2024-03-28T18:03:45Zhttps://upcommons.upc.edu/oai/requestoai:upcommons.upc.edu:2099.1/155262020-02-12T20:43:38Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Valls Solé, Gaspar
author
2009-01
El marc de referència general del projecte de màster realitzat és el disseny i anàlisi
d’una solució tecnològica completa per a la supervisió constant i de la salut i detecció
d'anomalies de les persones. Dins d’aquest marc de referència global, el treball específic
desenvolupat s’ha basat en validar el funcionament d’un sensor inercial per a la detecció
i supervisió de certes disfuncions que es produeixen de manera habitual en l'activitat
diària de moltes persones, principalment d’edat avançada, i que en qualsevol cas són
molt freqüents en la majoria de pacients que requereixen d’un cert grau de supervisió.
Una vegada validat el funcionament del mòdul sensor, aquesta tecnologia ha de
significar el primer pas per arribar a crear una arquitectura de supervisió de salut de les
persones que permeti disposar d’una informació constant i ubiqua del pacient.
Per al desenvolupament de la investigació s’ha utilitzat un sensor inercial Xsens MT-9
que disposa d’acceleròmetres, giròscops i magnetòmetres de 3 dimensions amb sortides
calibrades. A més, el sensor es subministra amb una sèrie de llibreries que ofereixen una
posada en marxa simplificada de l’experimentació i anàlisi de senyals.
S’ha optat per realitzar l'anàlisi experimental sobre MATLAB, la qual ha de permetre
realitzar la detecció i la supervisió d’una persona en temps real. Des del punt de vista de
la supervisió, els algorismes implementats analitzen la marxa i l’equilibri de l’individu
tot detectant, en un primer moment, els moviments laterals involuntaris i la despesa
energètica. Els moviments laterals involuntaris denoten canvis en l’estat de salut,
l’augment en el risc de caigudes o simplement són signe d’un deteriorament de la salut
de la persona. El seguiment de la despesa energètica permet observar l'augment o
reducció de l'activitat de la persona que vesteix el sensor. Aquesta supervisió ha de
generar una història per tal que pugui ser analitzada per un centre mèdic i permeti, a
partir d’ella, fer una futura diagnosi.
A més a més, des del punt de vista de la detecció, l’algorisme permet la detecció de
caigudes i, per tant, podria ser utilitzat per generar una alarma a serveis d’urgència
mèdica.
Una vegada validat aquest sistema, es disposa d’un punt de partida òptim per arribar a
construir, de manera escalable, una arquitectura de supervisió de la salut. Per una banda
es poden utilitzar més mòduls basats en els mateixos sistemes inercials que disposa el
sensor comercial, però implementant una comunicació sense fils, però també es poden
incorporar nous tipus de sensors, com ara d’electrocardiograma, de pressió arterial, etc.
D’altra banda el fet de disposar de més sensors pot facilitar amb el desenvolupament
dels algorismes corresponents, la detecció d’informació addicional, com ara, l’activitat
motriu de la persona, que a més, permet que la diagnosi es faci depenent de la tasca que
la persona realitza en cada moment. Finalment s’ha de construir l’arquitectura ubiqua, és a dir, que sense intervenció del
pacient, la informació captada i analitzada per l’arquitectura personal sigui transmesa en
temps real al món exterior: centres d’urgències en cas de detecció d’una situació
d’emergència o centres mèdics que recullin la informació històrica de l’individu.
http://hdl.handle.net/2099.1/15526
Àrees temàtiques de la UPC::Ciències de la salut::Medicina
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Enginyeria electrònica::Instrumentació i mesura::Sensors i actuadors
Supervisió i detecció d'activitats d'una persona mitjançant sensors inercials vestibles
oai:upcommons.upc.edu:2117/1674152022-12-25T01:08:00Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Bermeo Ayerbe, Miguel Ángel
author
2019-07-10
http://hdl.handle.net/2117/167415
ETSEIB-240.144758
Àrees temàtiques de la UPC::Energies::Eficiència energètica
Energy efficiency improvement of an industrial process test-bench using predictive control with model adaptation capabilities
oai:upcommons.upc.edu:2117/1655152020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Rios Cruellas, Daniel
author
2018-06-21
http://hdl.handle.net/2117/165515
ETSEIB-240.133867
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
In-contact drone flight navigation control.
oai:upcommons.upc.edu:2117/3729242022-09-20T06:08:22Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Mañas Pellejero, Pablo
author
2022-07-26
Aquesta memòria tracta sobre la realització d’una aplicació que serveix com a interfície d’usuari (HMI o “Human Machine Interface”) gràfica d’alta resolució i tàctil als missatges rebuts a través d’un bus CAN mitjançant l’ús d’un dels microcontroladors més moderns que existeixen al mercat, la placa de desenvolupament STM32H735G-DK la qual forma part de
la família STM32 de microcontroladors. També es dona informació sobre els diferents elements que formen part del projecte, és a dir, s’explica el funcionament del bus CAN, l’entorn de programació amb STM32 i, en definitiva, tots els aspectes relacionats amb el tema que es consideren importants. Cal dir que per aconseguir l’objectiu principal s’han fet servir les eines que el fabricant STMicroelectronics proporciona als seus usuaris: TouchGFX Designer que facilita el disseny d'una interfície amb l'usuari amb les seves plaques d'avaluació; STM32CubeIDE com a plataforma de desenvolupament, generació, compilació i depuració de codi en llenguatge C i C++; i STM32CubeMX per a la configuració de les freqüències dels rellotges interns, els pins i les interrupcions corresponents per al correcte funcionament de la comunicació a través del bus CAN. Com a resultat s’ha obtingut una aplicació funcional que permet a l’usuari rebre i enviar missatges a través del bus CAN i observar-los de manera gràfica a la pantalla d’una placa STM32H735G-DK. Com a conclusió, encara que algunes de les qualitats del programa (interaccions i disseny) podrien ser diferents, el programa final compleix els requisits que s’esperen
http://hdl.handle.net/2117/372924
ETSEIB-240.171408
Àrees temàtiques de la UPC::Enginyeria electrònica::Microelectrònica::Sistemes digitals programables
Desenvolupament de gràfics amb STM32 amb missatges CAN
oai:upcommons.upc.edu:2117/988962020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Martínez Hurtado, Iván
author
2016-09-05
This project proposes a reliable fault-tolerant MPC applied to DWTNs. The proposed approach the proposed approach allows to carry out a degradation analysis of the system in terms of performance and reliability. As a result of this analysis, the MPC controller design can be modified adapting the constraints such that the best achievable performance with some pre-stablished level of reliability will be achieved. The proposed approach will be tested on the Barcelona Water network.
http://hdl.handle.net/2117/98896
ETSEIB-240.119237
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Reliable Fault-Tolerant Model Predictive Control of Drinking Water Networks
oai:upcommons.upc.edu:2099.1/162722023-12-03T01:17:10Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Cappellino, Ezio
author
2009-09
El proyecto prev e el estudio y el desarrollo de un sistema de teleoperaci on
con control basado en realimentaci on h aptica.
Se ha desarrollado una aplicaci on en tiempo real que permite al usuario
controlar remotamente un Robot a trav es de un dispositivo maestro h aptico
Phantom Omni de 6 grados de libertad.
Una de las tareas principales de este proyecto ha sido el estudio exhaustivo
de la programaci on y control de una interfaz h aptica para poder utilizarlo
como master del sistema de teleoperaci on y poder realizar as un control de
impedancia basado sobre fuerzas calculadas a partir de la posici on del Robot.
El desarrollo del proyecto ha sido util tambi en para poder familiarizarse
con el entorno de los robots STAUBLI y con el funcionamiento de las
controladoras basadas en V+.
Adem as se ha desarrollado un retorno visual, realizado en OpenGL, para
poder guiar el robot de forma correcta sobretodo en la primera fase de
simulaci on.
Para demostrar y evaluar las capacidades del dispositivo h aptico ha sido
implementada una aplicaci on que permita al usuario realizar una tarea b asica
de manipulaci on que consiste en mover el elemento terminal del robot a
lo largo de un recorrido de complejidad elevada, excavado en una pieza de aluminio.
El sistema provee una realimentaci on de fuerza en tres grados de libertad
que ayuda al usuario a no colisionar con las paredes laterales del recorrido y
no permite mover el Phantom a una velocidad mayor que la del robot. Se han
realizado 3 pruebas con diferentes usuarios y medido el n umero de colisiones
realizadas como ndice de evaluaci on de la aplicaci on:
1. El usuario realiza el recorrido normal utilizando solo la realimentaci on
visual.
2. El usuario realiza el recorrido normal con el soporte de la realimentaci on
de fuerza.
3. El usuario realiza el recorrido con la realimentaci on de fuerza y la pieza
real es de una escala inferior respecto al modelo virtual percibido.
http://hdl.handle.net/2099.1/16272
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Informàtica::Robòtica
Implementaci ón de un sistema de teleoperaci ón con realimentaci ón H áptica
oai:upcommons.upc.edu:2099.1/162732020-02-12T20:43:38Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Gómez Zuluaga, Lady Johanna
author
2010-09
En el ámbito de la vivienda automatizada y hacia contextos de inteligencia ambiental, es necesario abordar la interacción entre la persona y los sistemas automáticos como parte integrante del diseño. Para tratar esta interacción se ha implementado un sistema interactivo de ayuda en la forma de un avatar dentro de una interfaz domótica que simula una vivienda automatizada. El sistema ofrece al usuario un soporte para la comprensión y manejo de la interfaz. Todo ello se fue desarrollando y mejorando gradualmente en función de los resultados de los estudios de usabilidad de las diferentes pruebas experimentales llevadas a cabo hasta conseguir implementarlo como sistema interactivo de soporte al usuario.
http://hdl.handle.net/2099.1/16273
Àrees temàtiques de la UPC::Edificació::Instal·lacions i acondicionament d'edificis
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Informàtica::Robòtica
Estudio del uso de una interfaz domótica con la incorporación de un sistema de ayuda interactivo
oai:upcommons.upc.edu:2117/1085532020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Capelo Ramon, Daniel
author
2017-09-13
Se desea iniciar el trabajo y estudio de algoritmos de computer vision sobre la plataforma robótica CASPER. Estos algoritmos deben de aunarse con el resto de componentes de procesado del robot.
http://hdl.handle.net/2117/108553
ETSEIB-240.127849
Àrees temàtiques de la UPC::Informàtica
Implementation of computer vision algorithms for the CASPER Robot
oai:upcommons.upc.edu:2117/1814472022-09-18T00:17:58Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Manzini, Enrico
author
2019-07-10
Linguistic differences between specialists and laymen still represent an obstacle for a successfull communication in technical environments. This is especially true in the medical domain where the linguistic gap between clinicians and patients is a considerable issue: from one side, diseases and symptoms must be described with a very specific vocabulary to avoid doubts and ambiguity; from the other side, it can not be expected for patients, that are the main source of information for an accurate diagnosis, to use the same technical jargon as physicians in order to describe their symptomatology. The main objective of this project is to investigate a possible solution to this issue using a deep learning approach to support the collection and description of all the traits of patients with rare disease in the Share4Rare network. Machine learning techniques will be used to develop a machine translation model that will be able to transform the input layman terms into specific medical concepts. In order to achieve this objective, the most common deep learning methods used in Natural Language Processing will be explained and analyzed, with a particular focus on word embedding techniques, convolutional neural networks and recurrent neural networks. Then, three models that combine these techniques will be proposed, trying to outline strengths and weak- nesses of each one. All the models will be created and tested with Python, a high-level, general- purpose programming language. The neural network architectures will be created using Keras, an open-source deep learning library for Python. The proposed models will be trained and tested using the lexicon from the Human Pheno- type Ontology, a formal ontology of human phenotypes with the aim of becoming the standard vocabulary for clinical databases. Terms in the Human Phenotype Ontology contain synonyms and descriptions of the phenotypes to which they refers that will be used as input for the different models. Results will be evaluated with cross-validation, and domain specific performance metrics will be adopted to carry out a specific analysis of the outcomes
http://hdl.handle.net/2117/181447
ETSEIB-240.145506
Àrees temàtiques de la UPC::Informàtica
Automatic translation between layman and HPO terms using machine learning algorithms
oai:upcommons.upc.edu:2117/3949082024-03-03T20:32:09Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Dalmases Mendoza, Albert
author
2023-09-20
Aquesta tesi presenta la implementació d’un marc de percepció robòtica avançada destinat a millorar la capacitat d’un robot per interpretar i interactuar amb el seu entorn. Donada la creixent presència de la robòtica en diversos sectors, hi ha una demanda cada vegada major de sistemes de percepció més intel·ligents i flexibles. Aquest treball satisfà aquesta demanda combinant les dades provinents de càmeres, implementant tècniques de detecció d’objectes i d’estimació de posició i orientació, i introduint capacitats de raonament intel·ligent. Utilitzant el “Robot Operating System” (ROS), el marc de percepció proposat combina mètodes basats en marcadors fiducials i en aprenentatge profund per a la detecció d’objectes i l’estimació de posicions i orientacions. Un aspecte original d’aquesta contribució resideix en la incorporació d’ontologies per estructurar i raonar sobre les dades percebudes. El resultat és un sistema de percepció robust i intel·ligent capaç de manejar escenaris complexos simulats i en el món real. Aquest mòdul de percepció proporciona funcionalitats que milloren la comprensió de l’entorn per part d’un robot, centrant-se principalment en tasques de manip- ulació robòtica. El treball presentat proporciona una base sòlida per a futures investigacions i ampliacions.
Esta tesis presenta la implementación de un marco de percepción robótica avanzada destinado a mejorar la capacidad de un robot para interpretar su entorno e interactuar con él. Dada la creciente presencia de la robótica en diversos sectores, hay una demanda cada vez mayor de sistemas de percepción más inteligentes y flexibles. Este trabajo satisface dicha demanda com- binando los datos proveninentes de cameras, implementando técnicas de detección de obje- tos y estimación de posición y orientación, e introduciendo capacidades de razonamiento in- teligente. Utilizando el ’Robot Operating System’ (ROS), el marco de percepción propuesto combina métodos basados en marcadores fiduciales y en aprendizaje profundo para la detec- ción de objetos y la estimación de posiciones y orientaciones. Un aspecto original de esta con- tribución reside en la incorporación de ontologías para estructurar y razonar sobre los datos percibidos. El resultado es un sistema de percepción robusto e inteligente capaz de manejar escenarios com- plejos simulados y en el mundo real. Este módulo de percepción proporciona funcionalidades que mejoran la comprensión del entorno por parte de un robot, centrándose principalmente en tareas de manipulación robótica. El trabajo presentado proporciona una base sólida para futuras investigaciones y ampliaciones
This thesis presents the implementation of an advanced robotic perception framework intended to enhance a robot’s ability to interpret and interact with its environment. Given the increased prevalence of robotics across various industries, there is a growing demand for more intelligent and flexible perception systems. This work meets that demand by combining visual sensor data, implementing object detection and pose estimation techniques, and introducing smart reason- ing capabilities. Utilizing the Robot Operating System (ROS), the proposed perception frame- work combines fiducial marker-based and deep learning-based methods for object detection and pose estimation. A unique aspect of this contribution lies in the incorporation of ontologies to structure and reason over the perceived data. The result is a robust, intelligent perception system capable of handling complex simulated and real-world scenarios. This perception module provides functionalities that enhance a robot’s understanding of its surroundings, with a primary focus on robotic manipulation tasks. The developments presented provide a solid foundation for further research and expansion
http://hdl.handle.net/2117/394908
ETSEIB-240.181148
Àrees temàtiques de la UPC::Informàtica::Robòtica
A smart and flexible perception framework for robotic manipulation
oai:upcommons.upc.edu:2117/1669322020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Espinosa Muñoz, Miller Stiven
author
2019-07-04
This project aims to design and implement a solution for a complex mobile manipulation
task using PAL Robotics’ robot TIAGo. The goal is to use the platform as an assistive robot,
making it capable of picking a chosen object from a certain location and moving it to another. More precisely the task in which the project will focus will be to pick a certain soda
can from a table with several of them on it and pour its contents into a glass placed on
another table for the user to have his/her drink . For that, all the perception, interaction,
planning and mobile capabilities of the robotic platform will be exploited in order to develop
a suitable and complete solution.
The generated solution is a modular method that can be launched as a whole process to
complete the entire challenge, but this solution also gives the possibility of using each
one of the modules independently. This way it is feasible to easily integrate them to other
processes in order to complete other similar tasks, making the packages more versatile and
adaptable.
The entire project has been divided in two parts. One focused on developing the packages
in charge of navigation and arm manipulation, carried out by Xavier Garcia Peroy. The
other one, focused on developing perception and task management part, and described in
this report.
For the perception part, some computer vision capabilities have been implemented using
TIAGo’s camera. These capabilities were added in order to get knowledge about the objects with which the robot needs to interact. In this case, these objects were cans of soda
which the robot needed to detect and, using some image processing steps, determine their
position in order to pick the can of soda chosen by the user.
On the other hand, for task management part, a solution based on behavior trees was
developed. This solution has been done in a modular way, removing a big part of the
complexity of executing each necessary task to achieve the goal of this project, and also
decreasing the necessity of going deep in programming in order to make changes and check
partial functionalities. That has been achieved with the use of the BehaviorTree library and
Groot, a visual tool that has allowed to create the task manager functionalities graphically.
Both solutions have been checked to work properly in order to achieve the established goal.
However, specially in perception part, some functionalities should still be improved in order
to increase the robustness of the method and decrease some limitations. Future work is
suggested in order to make these enhancements.
http://hdl.handle.net/2117/166932
ETSEIB-240.145022
Mobile manipulation with the TIAGo robot: perception and task manager
oai:upcommons.upc.edu:2117/3698662022-07-08T08:00:21Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Real Vial, Marta
author
real vial
author
2022-05-23
The navigation of Autonomous Underwater Vehicles (AUV) presents a series of challenges, being the location due to the inaccessibility of GPS signal one of them. A solution is the use of several surface vehicles equipped with hydrophones to locate and track an AUV that emits an acoustic signal. This Master’s Thesis addresses this solution with the objective to optimize the energy consumption of 2 Autonomous Surface Vehicles (ASV) while maintaining a preferred geometry of the formation to reduce the uncertainty of the position estimation of the AUV. To control these surface vehicles, 2 approaches of Reinforcement Learning (RL) have been implemented, Deep Deterministic Policy Gradient (DDPG) and Multi-Agent Deep Deter- ministic Policy Gradient (MADDPG). The aim of using these 2 algorithms is to test if the inclusion of the multi-agent element is necessary in this case. The RL algorithms control the linear and angular velocities of each robot. The environment used for the training of the robots is created specifically for this project using Python. A weighted sum of Gaussian functions is designed for the reward function, which contains all the elements related to the optimization of energy consumption and the formation of the AUV. To analyse several aspects of the final models of each robot a total of 4 different tests were done. These tests focus on the analysis of the distribution of the weights in the reward func- tion and the ability to adapt to difficult scenarios. The tests are done as simulations. The re- sults show that the ASVs trained with these implementations of RL modify their behaviour depending on the weight configuration and can adapt to more difficult scenarios depending on aspects of the training such as the noise applied to the actions. Moreover, the performance between the DDPG and MADDPG is compared. A discussion with similar works that have treated this problem is done using the final results.
http://hdl.handle.net/2117/369866
ETSEIB-240.164311
Àrees temàtiques de la UPC::Enginyeria electrònica
Àrees temàtiques de la UPC::Nàutica::Enginyeria naval
Simultaneous AUV navigation and tracking with satellite ASVs
oai:upcommons.upc.edu:2117/1806192020-05-11T04:55:12Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Doral Vall, Judith
author
2020-02-14
Nowadays, to ensure the quality of the joining technics that are used in the automotive manufacturing for instance welding, brazing, clamping, bolting and screwing; manual inspections are made. Those inspections are made with the aim of guarantee the correct condition of the joints. They are manual skillful processes that require the destruction of the product. Moreover, this tasks turn out to be heavy, slow-witted, toilsome, dangerous and inefficient in an economic point of view. Lately, ultrasonic technology has begun to be used to guarantee the quality of certain joining techniques. The field where is more developed is the one of welding. Nevertheless, these ultrasound devices aren’t able to substitute the destructive tests as it can’t manage with all the joining techniques and all the complex automotive shapes. Therefore, although the companies destroy fewer pieces; they still need a lot of time and people to inspect properly. In this project, an exhaustive study is carried out on the methodology used at the moment in an automotive factory and innovative solutions are proposed in order to improve this process and make it more efficient. For this reason, the objective will be to adapt the outdated process at the significant transformation regarding the way we produce products we’re in the midst of: Industry 4.0
http://hdl.handle.net/2117/180619
ETSEIB-240.147991
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Innovation in an automotive industry of the welding points quality tests
oai:upcommons.upc.edu:2117/1094262020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Riera Chiquito, Abrahan Alfredo
author
2017-09-13
This thesis is devoted to apply a Multi-Input Multi-Output (MIMO) controller to a
specific Steam Boiler Plant. The considered plant is based on the descriptions obtained
from the input/output data of a referenced steam boiler in the Abbot combined cycle
plant in Champaign, Illinois. The objective is to take all the useful input/output data
from the steam boiler according to its performance and capability in different operation
points in order to model the most accurate plant for control. The conceived case of study
is based in a modification of a model proposed by Pellegrinetti and Bentsman in 1996,
considering to be tested under a benchmark proposed by the Control Spanish Association
(CEA).
Initially, taking into account only the input and output data of the system, black box
modeling techniques were used to obtain different models of the plant. The first approach
was to obtain a transfer function model to apply a Internal Model Controller (IMC). However
the result was not as expected because the controller becomes considerable difficult
to tune given the big quantity of poles and zeros of the resulting IMC controller. Hence
this technique was dismissed.
On a second stage, it was obtained a model of the plant in state space representation
to apply a Linear-Quadratic Regulator (LQR) technique to understand how the system
behaves with this state space model design. Given that the description of the system in
this form was more accurate the obtained results were better for this type of controller
making it better suited to fulfill the needs of the plant.
This work covers all the steps followed to use the Internal Model Controller (IMC)
and the Linear-Quadratic Regulator (LQR) techniques to study the behavior of a steam
boiler system in an industrial environment. The obtained results are exposed and explained
with the aim of describing which one of the two used methods is better suited
for the control of the plant. Finally a budget and impact studies are presented to explain
which could be the resources needed in order to apply this type of controllers effectively
in a steam boiler plant, being able to extrapolate the obtained results to be applied to
other type of processes in the same sector (heat exchangers, distillation columns, etc.).
http://hdl.handle.net/2117/109426
ETSEIB-240.120073
Àrees temàtiques de la UPC::Energies
Multivariable control of a steam boiler
oai:upcommons.upc.edu:2117/3718272024-03-03T20:32:43Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Tezcan, Benjamin Harun
author
2022-07-15
Traditional machine learning methods amongst others segment an image into different regions on basis of pixel attributes. One of the most well known methods are clustering and thresholding algorithms. Convolutional neural networks are designed to imitate the human visual cortex by applying convolutional filters in form of layers to the input images. That way, deep features are extracted. Since this process is often referred to a “black box” behaviour because it is unknown what is going on inside the model, deep learning interpretability methods are introduced that highlight parts of an image that are important for the algorithm’s classification process. Experiments show that it is possible to establish links between segmented regions extracted by traditional methods and deep features extracted by CNNs. These results rely heavily on
the DL interpretability method used and the type of dataset. This extended TFM is carried out in cooperation with a research group from the AMRITA University – school of medicine in India, focusing on biomedical image processing and computer vision. The collaboration started in November 2021
http://hdl.handle.net/2117/371827
ETSEIB-240.166060
Àrees temàtiques de la UPC::Ciències de la salut
Establishing links between image segmentation and deep learning interpretability methods
oai:upcommons.upc.edu:2117/985202020-02-12T20:43:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ruiz Ramírez, Eduardo
author
2016-09-05
During the last years, new technologies approaches have helped to develop realistic robotic hands for prosthetic use. Even so, the strategies to control them (input signals, prediction algorithms) are still limiting a complete match between the robotic hand and the real hand movements and behaviors. On this thesis, two different input signals (FMG and sEMG) were evaluated. From this analysis characteristic properties from each kind of signal were obtained, related with wrist and hand movements. In this way two different learning methods were implemented for the first time on robotic hand research. The goal of these two methods was to combine both kind of input signals, supported by the feature analysis previously done, in order to improve the movements prediction performance. The methods’ performance were compared with the separate input signals methods, so the improvement could be measured. Both mixing methods presented better results than the single input signal ones. These results along with other considerations defined, could lead to a robotic hand performance improvement from different perspectives
http://hdl.handle.net/2117/98520
ETSEIB-240.120820
Àrees temàtiques de la UPC::Informàtica::Robòtica
Control of a hand prosthesis using mixed electromyography and pressure sensing
oai:upcommons.upc.edu:2117/3325082023-12-03T01:39:44Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Vila Perez, Alexandre
author
2020-07-30
The increasing demand on food leads to applying agriculture methods to increase the overall crop productivity. Most of the farmers decide to use chemical herbicides to get rid off the weeds that grow among the crop in order to avoid crop losses. Most of these chemical products cause an increase on the pollution on soil, on air and on water, damaging the flora, fauna and of course the humans. On the other hand, the concept of organic agriculture is recently being demanded by a lot of people that are conscious of the intrinsic dangers that cause the chemical herbicides used in agriculture. The drawbacks of the organic agriculture is that it takes time and effort, translated into additional costs for the farmer. That’s why farmers with large extension of crop fields are starting to try different tools in order to automate these tasks. Most of these tools are based on robotic systems or autonomous vehicles like Unmanned Aerial Vehicles (UAV) or Unmanned Ground Vehicles (UGV) that are able to perceive the current state of the field like crop health, total crop area, etc., and in some cases, these vehicles are also able to act over the field by performing actions like weeding or the process of getting rid of the weeds. Other methods that can be inside or not of the organic agriculture, are the precision agriculture techniques, mostly based on not only to cover the supervision of an area of crop, but the crop individually. The methods of mapping the position of each crop and to retrieve the health or the volume of a single crop, even to detect and remove any weed detected, are inside the precision agriculture concept. In precision agriculture pipeline, first of all it is necessary to detect what is desired to detect. If this is achieved, then the mapping and other geometric techniques can be performed properly. But the first part based on the detection stage is not as straightforward as it may seem. Classical machine learning techniques based on computer vision like localizing objects from the image, are based on image binarization by threshold and then on counting the blobs of the binary image. This method can only be applied if the physical conditions that are taken into account during the parameter tuning, are the same than the conditions taking place in each of the future situations. Concerning a real-world application, it has to be assumed that there will be changing physical conditions. That’s why it is necessary to use the deep learning techniques in order to handle or generalize the predictions or classifications when the conditions change or when unknown situations are present when the robotic system is supervising the crop. This project is focused on the weed localization or targeting, and it is one of many others that are taking place in Earth Rover Europe, S.L. The main project of the company is to develop a rover or a mobile robot able to supervise and map the crop individually and also to remove weed plants from the crop fields.
http://hdl.handle.net/2117/332508
ETSEIB-240.152845
Àrees temàtiques de la UPC::Enginyeria agroalimentària::Agricultura
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
Weed Detection in Crop Fields using Deep Learning Techniques
oai:upcommons.upc.edu:2117/1873312022-06-12T00:36:00Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Manjón Prado, Javier
author
2020-05-12
This project studies the aggravating problem of road traffic and congestion, and some of its solutions: autonomous cars and traffic simulation software. Unreal Engine is presented as a new option for simulating a traffic environment. An application is programmed that allows the user to create and edit roadtracks by clicking and dragging in the editor, along with autonomous vehicle models to populate them with. The simulator is tested using different scenarios and obtaining relevant data. Finally, the obtained conclusions are mentioned, along with ideas and guidelines for future development
http://hdl.handle.net/2117/187331
ETSEIB-240.146412
Àrees temàtiques de la UPC::Informàtica
Using Unreal Engine as an engineering tool for traffic simulation and analysis
oai:upcommons.upc.edu:2117/1701422019-12-11T08:15:06Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Centellas Gil, Victor
author
2018
Twitter has become the largest microblogging platform where users can interact
between each other expressing opinions, thoughts and feelings related to any topic
or source of news in a compressed 280 character message, called tweet. Hashtags
are popular keywords used to label these tweets according to its content. This
work tries to nd out if the usage of hashtags to label tweets with similar content
is accurate enough. To do so, tweets from di erent popular hashtags have been
retrieved and processed in order to have a dataset with a content as close to
reality as possible. Several embedding methods and learning algorithms have been
studied to classify tweets from di erent hashtags based on the content. Results
showed that the best performance is achieved when using the Tf-idf embedding
method and support vectors machine. The learning algorithm obtained a precision
around 90% for classi cation on 10 classes and above 70% when dealing with 100
classes trained on datasets of only 13680 and 143067 samples respectively. The
results also indicated that BoW and Tf-idf methods outperformed other state of
the art methods for other natural language processing tasks, such as GloVe or
Word2Vec.
http://hdl.handle.net/2117/170142
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Telemàtica i xarxes d'ordinadors
Word Level Approach for Tweets Classification based on its Content
oai:upcommons.upc.edu:2117/3332722021-12-16T11:45:46Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Prieto Francia, Dunia
author
2020-07-15
Recently, industrialized countries carry out a technological race towards autonomous driving. The current commercial driver assistant systems are limited to aiding the driver while the driver is responsible for overall control of the vehicle or some autonomous driving functions under certain driving conditions. There are still many challenges to be solved to allow autonomous vehicles to emerge into the market. One of the key challenges for the development of fully autonomous vehicles is localizing the vehicle in known, unknown, or uncertain environments. Most of the existing works in the field of mobile robotics rely on a non-linear kinematic motion model under the assumption that the velocity is low. However, in autonomous vehicle applications where the velocity is higher, a dynamic model is required. On the other hand, conventional solutions are based on Extended Kalman filter that performs a linearization around the estimated trajectory. Within this context, this master thesis has considered a non-linear kinematic/dynamic motion model. Furthermore, the use of an LPV Kalman filter for estimating the vehicle dynamic state has been explored. Due to the philosophy underlying the design procedure, computational complexity is reduced and strong non-linearities are easier to handle. Kinematic states are obtained by means of numerical integration. The proposed approach has been implemented and integrated in an SLAM Toolbox for Matlab. Simulations results validate the developed localization technique. However, the results still show space for improvement. In particular, a better integration between the kinematic and dynamic parts should be developed. This can be achieved by designing an observer that integrates both the dynamic and kinematic parts. Another extension that is planned is the inclusion of landmarks in the LPV Kalman observer and the map building capability. Reaching autonomous driving it just a matter of time!!!
http://hdl.handle.net/2117/333272
ETSEIB-240.148967
Àrees temàtiques de la UPC::Enginyeria mecànica::Disseny i construcció de vehicles
Autonomous vehicle localization using state estimation techniques
oai:upcommons.upc.edu:2117/868022020-02-12T20:43:39Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Marco Valle, Alonso
author
2015-10
Robotic setups often need fine-tuned controller parameters both at low- and task-levels. Finding
an appropriate set of parameters through simplistic protocols, such as manual tuning or
grid search, can be highly time-consuming. This thesis proposes an automatic controller tuning
framework based on linear optimal control combined with Bayesian optimization. With
this framework, an initial set of controller gains is automatically improved according to the
performance observed in experiments on the physical plant.
In the tuning scenario that we propose, we assume we can measure the performance of the
control system in experiments through an appropriate cost. However, we only allow a limited
number of experimental evaluations (e.g. due to their intrinsic monetary cost or effort). The
goal is to globally explore a given range of controller parameters in an efficient way, and return
the best known controller at the end of this exploration.
At each iteration, a new controller is generated and tested on a closed-loop experiment in
the real plant. Then, the recorded data is used to evaluate the system performance using a
quadratic cost. We re-iterate in order to solve a global optimization problem, whose goal is to
learn most about the location of the global minimum from the limited number of experiments.
We use the Linear Quadratic Regulator (LQR) formulation as a standard way to compute
optimalmultivariate controllers given a linear plant model and quadratic cost. We parametrize
the LQR weights in order to obtain controllers with appropriate robustness guarantees.
The underlying Bayesian optimization algorithm is Entropy Search (ES), which represents
the latent objective as a Gaussian process and constructs an explicit belief over the location of
the objective minimum. This method maximizes the information gain from each experimental
evaluation. Thus, this framework shall yield improved controllers with fewer evaluations
compared to alternative approaches.
A seven-degree-of-freedomrobot arm balancing an inverted pole is used as the experimental
demonstrator. Results of a two-dimensional tuning problem are shown in two different
contexts: in the first setting, a wrong linear model is used to compute a nominal controller,
which destabilizes the actual plant. The automatic tuning framework is still able to find a stabilizing
controller after a few iterations. In the second setting, a fairly good linearmodel is used
to compute a nominal controller. Even then, the framework can still improve the initial performance
by about 30%. In addition, successful results on a four-dimensional tuning problem
indicate that the method can scale to higher dimensions.
Themain and novel contribution of this thesis is the development of an automatic controller
tuning framework combining ES with LQR tuning. Albeit ES has been tested on simulated
numerical optimization problems before, this work is the first to employ the algorithm for
controller tuning and apply it on a physical robot platform. Bayesian optimization has recently
gained a lot of interest in the research community as a principled way for global optimization
of noisy, black-box functions using probabilistic methods. Thus, this work is an important
contribution towardmaking thesemethods available for automatic controller tuning for robots.
In conclusion, we demonstrate in experiments on a complex robotic platform that Bayesian
optimization is useful for automatic controller tuning. Applying Gaussian process optimization
for controller tuning is an emerging novel area. The promising results of this work open
up many interesting research directions for the future.
In future work, we aim at scaling this framework further than this problem to higher dimensional
systems such as balancing of a full humanoid robot for which a rough linear model
can be obtained. In addition, comparing ES with other global optimizers from the literature
may be of interest. Investigating safety considerations such as avoiding unstable controllers
during the search is also part of future research.
http://hdl.handle.net/2117/86802
Àrees temàtiques de la UPC::Informàtica::Robòtica
Gaussian process optimization for self-tuning control
oai:upcommons.upc.edu:2117/855802022-05-22T00:51:46Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Rubio Romano, Antonio
author
2015-10
The present Master Thesis describes a new Pose Estimation method based on Convolutional
Neural Networks (CNN). This method divides the three-dimensional space in several regions
and, given an input image, returns the region where the camera is located.
The first step is to create synthetic images of the object simulating a camera located at
di↵erent points around it. The CNN is pre-trained with these thousands of synthetic images
of the object model.
Then, we compute the pose of the object in hundreds of real images, and apply transfer
learning with these labeled real images over the existing CNN, in order to refine the weights
of the neurons and improve the network behaviour against real input images.
Along with this deep learning approach, other techniques have been used trying to improve
the quality of the results, such as the classical sliding window or a more recent class-generic
object detector called objectness.
It is tested with a 2D-model in order to ease the labeling process of the real images.
This document outlines all the steps followed to create and test the method, and finally
compares it against a state-of-the-art method at di↵erent scales and levels of blurring.
http://hdl.handle.net/2117/85580
Àrees temàtiques de la UPC::Informàtica
3D pose estimation using convolutional neural networks
oai:upcommons.upc.edu:2117/1174512020-02-12T20:43:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Spanos, Konstantinos
author
2018-04-27
This work proposes a solution for the longitudinal and lateral control problem of urban autonomous vehicles using a gain scheduling Takagi-Sugeno (TS) control approach. Using the kinematic and dynamic vehicle models, a TS representation is adopted and a cascade control methodology is proposed for controlling both vehicle behaviours. In particular, for the control design, the use of both models separately will lead to solve two TS LMI-LQR problems. Furthermore, to achieve the desired levels of performance, an approach based on cascade design of the the kinematic and dynamic controllers has been proposed. This cascade control scheme is based on the idea that the dynamic closed loop behaviour is designed to be faster than the kinematic closed loop one. The obtained gain scheduling TS control approach, jointly with a trajectory generation module, has presented suitable results in a simulated city driving scenario.
http://hdl.handle.net/2117/117451
ETSEIB-240.133005
Àrees temàtiques de la UPC::Enginyeria mecànica
Takagi-Sugeno Fault Tolerant Control of an Autonomous Vehicle
oai:upcommons.upc.edu:2117/3719602024-03-03T20:33:50Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Li, Zhaoqi
author
2022-07-18
In nonlinear control teaching, traditional interactive tools such as Python and Sysquake based tend to be less flexible and single-function. Meanwhile, these tools are only adapted to restricted platforms and therefore, are dramatically limited for practical applications. In this project, a more flexible interactive tool has been designed and developed using MATLAB for the purpose of engineering teaching, offering a more general platform in order to do the auxiliary teaching in a more interesting way. Describing function method, an approximate extension of the frequency response method to nonlinear systems, is represented in MATLAB App designer in this project, which is important to predict the limit cycles. In addition, diagrams of certain nonlinearities, as well as the corresponding input and output signals, are shown intuitively in real-time through this interactive tool. In the meanwhile, validations from the Simulink block in MATLAB have been realized to verify the correctness of the results obtained in the interactive tool in this project. Some users who are doing the master’s degree in Automatic Control and Robotics at Universitat Politècnica de Catalunya have tested this tools, making them better understand the nonlinear theory behind and observe directly how parameters affect the other part in an interface
http://hdl.handle.net/2117/371960
ETSEIB-240.169902
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Matlab interactive application for nonlinearity
oai:upcommons.upc.edu:2117/3568032021-11-19T17:40:32Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Arabaci, Fatma Nur
author
2021-10-11
This thesis proposes a framework for configuring Task Planning Problems flexibly in an automatic manner using two main modules which are the Perception Module and the Reasoning Module. In order to automatize the overall process, initially, a knowledge layer is generated manually, in which the information regarding the environment is stored using ontologies, whereas the environmental state where the task is taking place is observed with the help of the Perception Module. The knowledge layer is then reasoned within the Reasoner Module in order to automatically configure task planning problems by filling Planning Domain Definition Language (PDDL) [1] files. During this reasoning process, the information retrieved from the Perception Module is used. In this paper, both of these modules mentioned above are explained in detail before providing the results separately for each module. Then, in addition to individual results, a scenario is created within a lab environment to test the overall system including both modules. Furthermore, alternative areas where the Reasoning Module implementation can be benefited from is also discussed
http://hdl.handle.net/2117/356803
ETSEIB-240.161282
Àrees temàtiques de la UPC::Enginyeria electrònica
Perception and reasoning for the automatic configuration of task and motion planning problems
oai:upcommons.upc.edu:2117/970032020-02-12T20:43:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ruiz Camps, Claudi
author
2016-04-15
Not always science and companies share the same objectives, however a company’s need can be sometimes satisfied by applying science. Two of the common problems of a company that tries to work with data in the field of artificial intelligence are firstly how to get the data itself and secondly how to label it. This thesis presents a real case where
a company has to acquire data, label it and then create the best model that fits the data. After many experiments this thesis shows how a small number of training inputs and an Arachnid model, that combines 45 specialized CNNs and a classifier that finds the pattern behind the output of those CNNs, can improve the test accuracy of LeNet-5 from 93.85% to 99,00% when classifying 10 different classes of optical characters with a concrete dataset. 1.000 single characters were extracted randomly from around 10.000 images of receipts and 800 of them were used for the training and 200 for the test. The approach of this thesis is focused on a real and concrete problem of a company, trying to find the best solution by using science and at the same time taking into account the company’s need.
http://hdl.handle.net/2117/97003
ETSEIB-240.111785
Àrees temàtiques de la UPC::Informàtica
Segmentation, labeling and optical character recognition applied on receipt images
oai:upcommons.upc.edu:2117/987332022-08-21T01:01:15Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Aguilar Plazaola, José Agustín
author
2016-09-05
This project presents the application of a visual servo control to an industrial human-like robot, both using a simulation environment and in a real platform. In a visual servo scheme, the control loop is closed by a vision sensor, usually a camera (or more than one, in the stereo approach). The camera acquires the image of a defined target and, a control algorithm calculates the relative pose of robot-target and then, continuously sends commands to the robot in order to position it as required. The pose calculation and controller algorithms have been written in C++. The work has been carried out through a sequence of stages that are presented in this document. The first part goes through the basic theoretical ideas that support the design of the visual servo. It is composed of three main areas: computer vision, which deals mostly with the implementation of the vision sensor; robot kinematics, which allows define the equations that describe time evolution of the robot position, orientation, speed and joints values; and finally, the merge of both areas, the visual servo itself that makes up the control loop. A next section explains how the different tools and frameworks have been used to implement the control loop. Some of these tools are manufacturer proprietary programs, others are open source. There is a detailed description of how the simulation environment is set, the content of each of the blocks in the control loop and a basic explanation of the manufacturers program.
The results show how the robot (simulated and real) converges to the relative set point pose and is also able to track changes in the position and orientation of the target.
http://hdl.handle.net/2117/98733
ETSEIB-240.118927
Àrees temàtiques de la UPC::Informàtica::Robòtica
Implementation of a Visual Servo Control in a Bi-Manual Collaborative Robot
oai:upcommons.upc.edu:2099.1/237052020-02-12T20:43:39Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Li, Mingming
author
2014-06
This thesis firstly proposes a new approach of modelling an ethanol steam reformer (ESR) for
producing pure hydrogen. Hydrogen has obvious benefits as an alternative for feeding the proton
exchange membrane fuel cells (PEMFCs) to produce electricity. However, an important drawback
is that the hydrogen distribution and storage have high cost. So the ESR is regarded as a
way to overcome these difficulties. Ethanol is currently considered as a promising energy source
under the research for its renewable and easy-access. The modelled reformer has been divided
into two stages in order to miximize the output of pure hydrogen. Specifically, the hydrogen
is generated in the first stage by steam reforming (SR) over the Cobalt-based catalysts. The
second stage is used to purify the hydrogen from the mixture gas through a Pd-Ag membrane
separation. Unlike the conventional way, the two stages are carried out in a same module, since
only one heating device is required. The main idea of the modelling approach is to develop a
one-dimensional, pseudo-homogeneous and non-steady-state model based on the mass balance.
Finite difference method is applied to transform the partial differential equations (PDEs) into
ordinary differential equations (ODEs).
Furthermore, the control of the ESR system is also discussed in this thesis. Some Model
predictive control (MPC) strategies have been designed for the linear and non-linear models
of the ESR. The mathematical model of the ESR by using the proposed approach is used for
the design of a MPC controller. The first part of this thesis has been devoted to obtain the
non-linear mathematical model with the assumptions of isothermal and isobaric conditions and
perform simulations. The numerical linearisation of the mathematical model towards a controloriented
model has been performed as well as the method of model-order reduction with Hankel
norm approximation and temporal discretization with zero-order hold, which will be used as
the prediction model of the MPC. The control objectives of the ESR system are to maximize
the hydrogen yield and minimize the ethanol flow rate, which result in a multi-objective problem.
Open-loop and close-loop simulations are developed to highlight the proposed controller
results. The analysis has taken into account relevant indices such as the ethanol conversion,input smoothness and output error.
http://hdl.handle.net/2099.1/23705
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Energies::Tecnologia energètica
Model Predictive Control for Ethanol Steam Reformers
oai:upcommons.upc.edu:2099.1/237142020-02-12T20:43:39Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Wang, Ye
author
2014-06
This thesis is devoted to developing a robust Model Predictive Control (MPC) strategy based
on Gaussian Processes (GP), especially for Drinking Water Networks (DWN). Nowadays there
are many different MPC strategies developed for DWN, such as certain-equivalent MPC (CEMPC)
and chance-constrained MPC (CC-MPC). The general control objectives for DWN are
economic by managing the water supply to minimise water production and transport costs, all
the tanks running in safe ways with their limitations and reducing the undesired abrupt changes
by minimising their slew rate and obtaining smooth signals. For the deterministic system model,
the control objectives are elementary fulfilled. But the main challenge for DWN is to propagate
and incorporate exogenous and endogenous uncertainties to MPC closed loop over the prediction
horizon. Considering the control-oriented model of the DWN, the MPC controller design
is hereby divided into two parts: system disturbances forecasting and the robust MPC controller
design. Case studies based on Barcelona DWN have been executed to verify the performance
of proposed methodologies.
The first part of this thesis leads to forecast system disturbances by using GP. In a DWN system,
system disturbances come mainly water demands associated to consumer sectors. Hence,
it is necessary to model each water demand and forecast the water demand in a short term that
covers the MPC prediction horizon. GP regression is regarded as one of state-of-the-art regression
methods able to select model parameters by using Bayesian Inference theory with a
collection of past data. Besides, it is believed that the GP regression method has a difficult for
the multiple-step ahead forecasting. Hence, the Double-seasonal Holt-winters method is used
for forecasting the expected disturbances while the stochastic disturbances are forecasted by using
GP. Finally, the desired forecasting results are a set of Gaussian distributions over the MPC
prediction horizon.
The second part of this thesis is to incorporate the forecasting results from GP within MPC
closed loop. This MPC strategy based on GP is named GP-MPC. Using the given system model,
the deterministic state evolutions can be obtained while the uncertainty of state propagation over a given prediction horizon can be also achieved though the linear approximation of GP. Therefore,
the worst-case state evolutions over the MPC prediction horizon can also be determined
in the MPC cost function and constraints. The desired performance of applying GP-MPC in
the closed-loop system is that the system has more safety than the CE-MPC and meanwhile it
probably brings more expenses.
Comparisons of GP-MPC and previous developed approaches are carried out by a case study
of the three-tank system inspired in the Barcelona DWN. A set of key performance indicators are
defined to compare performances of different MPC strategies. Finally, through the simulation
results, the GP-MPC has the similar performance as the CC-MPC, both of which have much
more expenses than the CE-MPC. As a result of considering the uncertainties inside the system,
more expenses is necessary to maintain the safety of the whole system. Hence, the GP-MPC
is more advanced. Moreover, the proposed GP-MPC is required to be tested with the whole
DWN and using the real data from a DWN system. So the future works of this thesis have been
outlined.
http://hdl.handle.net/2099.1/23714
Àrees temàtiques de la UPC::Enginyeria química::Indústria dels processos químics
Model Predictive Control for DrinkingWater Networks based on Gaussian Processes
oai:upcommons.upc.edu:2117/1247422020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Maillot, Stéphane Bertrand
author
2018-09-17
This thesis presents how to register point clouds from RGB-D sensors to perform real-time 3D reconstruction of an indoor scene with the aim to build a representation of the environment useful for robot planning. Several common techniques have been explored and compared to determine benefits and drawbacks of each method for our setup. A more specific method was then developed based on this knowledge and has been compared with other methods in terms of rotation and translation errors. Finally these results are used to evaluate the method accuracy and determine in which conditions these results can be replicated
http://hdl.handle.net/2117/124742
ETSEIB-240.134052
Àrees temàtiques de la UPC::Informàtica
Point Cloud Data Fusion for Real -Time Applications
oai:upcommons.upc.edu:2117/1677122020-03-02T01:30:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Gómez Pomar, Juan Carlos
author
2019-07-23
http://hdl.handle.net/2117/167712
ETSEIB-240.145137
Àrees temàtiques de la UPC::Enginyeria mecànica::Disseny i construcció de vehicles
PEM fuel cell system modelling and control for automotive applications
oai:upcommons.upc.edu:2117/3953112023-12-03T02:28:47Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Flaqué Borrellas, Ferran
author
2023-10-17
Aquest projecte parteix de la idea d'emprar sensors impresos en 3d per tal de comandar una plataforma d'ajuda a la mobilitat (cadira de rodes). Els sensors han estat dissenyats en altres projectes i impresos en l'espai de fabricació digital de l'ETSEIB (Escola Tècnica Superior d'Enginyeria Industrial de Barcelona). Principalment, s'han utilitzat dos tipus de sensors, un que d'ara endavant s'anomenarà de tipus cinta i que consta bàsicament d'una cinta plàstica amb un component resistiu a l'interior i un de tipus joystick que conté 4 elements resistius (un per cada direcció). El principi de funcionament dels sensors es basa en la variació de la resistència elèctrica a conseqüència del desplaçament exercit sobre aquests. Per tal de poder extreure un senyal de control a partir d'aquests sensors, doncs, s'ha cercat la millor manera d'obtenir un senyal generat per una resistència variable (la del sensor) i s'ha considerat més útil mesurar el voltatge emprant un divisor de tensió o un pont de Wheatstone pel fet que només ens interessa obtenir un senyal amb el qual controlar la cadira i no mesurar la resistència en si. En aquesta fase, s'ha cregut necessari amplificar el senyal, ja que la variació entre els voltatges mínims i màxims llegits era reduïda i no s'aprofitava el rang de lectura dels convertidors analògic a digital (ADC). A més a causa del soroll s'ha cregut convenient aplicar algun mètode de filtratge del senyal. Un cop es disposava d'un mètode per llegir el senyal amb menys soroll, s'ha observat comportaments anòmals (caigudes de resistència a l'exercir una força de tracció sobre el sensor) i ha calgut fer assajos per poder modelar el comportament del sensor en funció del desplaçament exercit sobre aquest. Un cop s'ha establert el comportament del sensor a traccions, emprant la lectura d'aquest s'ha creat un senyal de control. Per tal de poder provar el funcionament del senyal de control generat pel sensor ha calgut simular la cadira, ja que aquesta no ha estat construïda encara (s'ha fet una proposta del disseny complet del circuit electrònic que s'haurà de fer servir un cop es materialitzi la cadira). La simulació s'ha fet partint d'un dibuix CAD de la cadira i fent servir Gazebo (que ve inclòs en la llibreria de ROS) com a simulador. Finalment, s'ha dut a terme l'anàlisi econòmica i ambiental del projecte
Este proyecto parte de la idea de usar sensores imprimidos en 3d con el objetivo de controlar una plataforma de ayuda a la movilidad (silla de ruedas). Los sensores han sido diseñados en otros proyectos e imprimidos en el espacio de fabricación digital de la ETSEIB. Principalmente, se han usado dos tipos de sensores, uno que en adelante se llamará de tipo cinta y que consta básicamente de una cinta plástica con un componente resistivo en su interior y uno de tipo joystick que contiene 4 elementos resistivos (uno para cada dirección). El principio de funcionamiento de los sensores se basa en la variación de resistencia eléctrica como consecuencia del desplazamiento ejercido sobre estos. Para poder extraer una señal de control a partir de estos sensores se ha buscado la mejor manera de obtener dicha señal a partir de una resistencia variable (la del sensor), se ha considerado que lo más útil es leer el voltaje usando un divisor de tensión o un puente de Wheatstone porque solo nos interesa poder generar la señal de control y no leer la resistencia en sí. En esta fase, se ha considerado necesario amplificar la señal debido a la baja variación del nivel de voltaje máximo y mínimo, ya que no se aprovechaba el rango de lectura de los diferentes ADC. A demás, debido al ruido, se ha creído conveniente aplicar algún método de filtrado de la señal. Una vez se disponía de un método para leer la señal con menor ruido, se ha observado comportamientos anómalos (caídas de resistencia al ejercer una fuerza de tracción sobre el sensor) i ha sido necesario realizar diversas pruebas para modelar el comportamiento del sensor en función del desplazamiento de este. Una vez establecido el comportamiento del sensor a tracciones, usando la lectura de este se ha creado una señal de control. Para probar el funcionamiento de esta señal de control, ha sido necesario simular la silla, ya que aún no ha sido construida. La simulación se ha hecho usando un dibujo CAD de la silla y usando Gazebo (que está incluido en la librería de ROS) como simulador. Finalmente, se ha realizado el análisis económico y ambiental del proyecto
This project is based on the idea of using 3D printed sensors with the aim of controlling a mobility aidance platform (wheelchair). The sensors have been designed in other projects and printed in the digital manufacturing space of the ETSEIB. Mainly, two types of sensors have been used, one that will from now on be called tape type and basically consists of a plastic tape with a resistive component inside, and a joystick type that contains 4 resistive elements (one for each direction). The principle of operation of the sensors is based on the variation of electrical resistance because of the displacement applied on them. To extract a control signal from these sensors, there has been some research on the different methods used to read variable resistances. It has been considered that the most useful thing is to read the voltage using a voltage divider or a Wheatstone bridge because we are only interested in being able to generate the control signal and not to read the resistance itself. In this phase, it has been considered necessary to amplify the signal due to the low variation of the voltage level as the reading range of the different ADCs was not fully used. In addition, due to noise, it has been considered convenient to apply some methods to filter the signal. Once achieved a reading with low noise, anomalous behaviours (drops in resistance when exerting a traction force on the sensor) have been observed and it has been considered necessary to perform various tests to model the behaviour of the sensor depending on the displacement applied. Once the behaviour has been established, using the reading of the sensors, a control signal has been created. To test the operation of this control signal, it has been necessary to simulate the chair, since it has not yet been built. The simulation has been done using a CAD sketch of the chair and using Gazebo (which is included in the ROS library) as a simulator. Finally, the economic and environmental analysis of the project has been carried out
http://hdl.handle.net/2117/395311
ETSEIB-240.178363
Àrees temàtiques de la UPC::Enginyeria electrònica::Instrumentació i mesura::Sensors i actuadors
Desenvolupament virtual i en context d'una arquitectura de sensors per comandar una plataforma d'ajuda a la mobilitat
oai:upcommons.upc.edu:2117/1105122020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Coll Pujals, Dídac
author
2017-09-13
The huge amount of people who suffer from some form of disability implies that more research has to be
done in order to develop efficient robotic technologies for them. Yet, research costs can not be neglected.
In this matter, software simulators play a significant role as an alternative approach to minimize such expenses.
This project aims to make use of OpenSim API and Simbody API, two C++ libraries in order to prove the
feasibility of coupling and simulating an exoskeleton model with a musculoskeletal model. Due to the few
information related about how to get it, this thesis embraces many others sub-goals strictly related to the
achievement of the primary purpose.
The first sub-objective has been to learn OpenSim and Simbody API. As solution proposed in this project
has been to create a double pendulum model in OpenSim.
The second one has been design a model of an exoskeleton. As a result, it has been implemented in
OpenSim, and a controller has been designed to test it.
The successful achievement of the previous sub-objectives have allowed to create the simulation of the
exoskeleton coupled to the musculoskeletal model and simulate it in OpenSim.
Last but not least, the documentation throughout the project in conjunction with the successful achievement
of the proposed goals has made possible the creation of this thesis which indeed aims to keep working with
the exoskeleton coupled to musculoskeletal model or even become a reference in future robotic orthosis
projects build in OpenSim and Simbody APIs.
http://hdl.handle.net/2117/110512
ETSEIB-240.127769
Àrees temàtiques de la UPC::Enginyeria biomèdica
Simulation of the assistance of an exoskeleton on lower limbs joints using Opensim
oai:upcommons.upc.edu:2117/1829912021-12-16T11:45:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Zornoza Uña, Alvaro
author
2019-10-07
Currently, there is a transition in industrial processes towards Industry 4.0. Manufacturing industries face several challenges, but the main goal is always to maximise asset availability. Today, poor maintenance strategies can reduce a plant’s overall productive capacity. Time-based preventive maintenance has proven to be an ineffective maintenance technique for most equipment components. Indeed, predictive maintenance is one of the main factors that will drive the fourth industrial revolution, making factories more efficient and productive, reducing unplanned downtime, improving reliability, and saving maintenance costs. Although predictive maintenance is a technology with promising results, its practical applicability is currently limited by the complex access to data from industrial equipment. The present project, first of all, aims for finding and developing a proper use case, despite not being in an industrial environment and not having access to real field devices. Electromechanical relays are used in many of today’s machines where control of high voltages and currents is intended. They are indeed important components in the context of automation, so it was found that the performance of accelerated life-test of electromechanical relays was a good use-case for predictive maintenance. This kind of life-test usually require dedicated hardware but in this project, it was proposed to conduct the experiments using standard automation components. SIMATIC Fail-Safe controllers and F-DI modules and its safety-related capabilities, even if they are not intended equipment for this purpose, were used to conduct an accelerated life test of general-purpose electromechanical relays based on the concept of discrepancy analysis. The present project proposes a solution based on a SIMATIC system with a CPU S7-1500 and F-IO modules, a setup with electromechanical relays to be tested and a single board computer. In order to enable the experiments, the vii firmware of the F-DI module was modified to monitor the operation of the relays throughout the experiment and stream the data to the single board computer for further data analysis. The experiments aim to observe whether it is possible to detect any change before the relay breaks. It was found that the time resolution achieved by the original firmware was not enough for the requirements of the project. To solve this issue, a new discrepancy algorithm was developed. Due to time constraints, it was only possible to perform a few experiments. Moreover, it was not possible to identify clear patterns that are common to all the experiments. Although the results of the experiments were not as satisfactory as could be expected, this work presents a proven solution that could be applied to other cases of use in the future. Moreover, this research could be extended by performing more experiments which allow concluding whether the performance of accelerated life tests of relays is a feasible concept
http://hdl.handle.net/2117/182991
ETSEIB-240.143800
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Enabling fault prediction of electromechanical relays by incorporating accelerated life tests using Siemens SIMATIC safety controllers
oai:upcommons.upc.edu:2117/3770142024-01-28T02:36:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Pedrosa Alias, Javier
author
2022-09-12
The development of this thesis aims to design a new approach for solving production planning and scheduling in the process industries in such a way to be adaptable to any manufacturing plant, the description of which would have to be previously provided together with a series of ordered jobs. The planning and scheduling solving is concerned with the allocation over time of scarce resources between competing activities to meet a given set of requirements with an efficient organization. But, things get complicated as larger the scale of the problem is, i.e. as more resources, activities and requirements are involved. That is why the orientation of the work is focused on an innovative method in the style of Artificial Intelligence, by means of an automated process seeking to converge to a predefined objective. Although this research object has been studied since the middle of the last century, major breakthroughs were not achieved until the emergence of high-performance computing technologies; since by nature these are combinatorial problems which, the larger the scale, the more exploration they require to find some optimal. In addition, most of the last years related articles has been focused on solution approaches based on mathematical programming techniques, and it is important to note that there are other solution methods for dealing with this kind of problems. These methods can be used either as alternative methods, or as methods that can be combined with mathematical programming models, like the one proposed in this document
http://hdl.handle.net/2117/377014
ETSEIB-240.170439
Àrees temàtiques de la UPC::Economia i organització d'empreses::Direcció d'operacions::Modelització de transports i logística
General advanced job shop scheduling approach
oai:upcommons.upc.edu:2117/986622021-11-07T02:25:27Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Pumarola Peris, Albert
author
2016-06-23
In this Master Thesis one of the most common problems related to face detection is presented: fast and accurate unconstrained face detection. To deal with this problem a new general learning method is presented. The proposed method introduces a set of upgrades and modifications on key concepts and ideas of Decision Trees, AdaBoost and Soft Cascade learning techniques. Firstly, a new variation of Decision Trees with quadratic thresholds able to maximize the margin distance between classes is introduced. Considering a training set independent of face orientation and viewpoints information, the proposed algorithm is able to learn a combination of features to cluster faces under unconstrained face position and orientation. Next, a new definition of the Soft Cascade thresholds training principles is provided. Hence, this modification leads to a better formulation of the loss function associated to the AdaBoost algorithm. The trained face detector has been tested over the Face Detection Data Set and Benchmark (FDDB) and compared against the current state of the art classifiers. The obtained results show that the proposed face detector (i) is able to detect faces with unconstrained position, and (ii) it works faster than the current state of the art methods
http://hdl.handle.net/2117/98662
ETSEIB-240.119369
Àrees temàtiques de la UPC::Informàtica
Optimization of common computer vision algorithms : beating OpenCV face detector
oai:upcommons.upc.edu:2117/997342019-03-06T06:47:34Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
García Rodríguez, Carlos
author
2015
The ADREAM1 project (French acronym for Embedded Reconfigurable Dynamic Autonomous and Mobile Architectures) is a research program focused on Smart Grids, as well as on Ambient Physical Cyber Systems. The domains of integrated systems, robotics, and networks, have progressed considerably during recent years. That is, emerging systems that interact in meaningful ways with their environment and with constrains that are highly variable and dynamic. These systems are heterogeneous, integrated, and mobile, while possessing detection and action functions. Thus, they must respond to operational constraints in terms of time and energy by being adaptive and resilient.
http://hdl.handle.net/2117/99734
Àrees temàtiques de la UPC::Informàtica::Robòtica
The Power Consumption of the Geothermal Plant in ADREAM, a LAAS-CNRS Building. Modeling with Matlab-Simulink.
oai:upcommons.upc.edu:2117/3335602023-12-03T02:52:28Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Pieras Morell, Tomàs Josep
author
2020-07-30
Robotic Minimally Invasive Surgery is one of the surgical fields in which robotics is most extended. Minimally Invasive Surgery is used in fetal surgery for the correction of TTTS, among others pathologies. This technique has great complexity and long learning curves. To improve this surgery a robot-assisted TTTS surgery-oriented teleoperation platform was developed by the robotic division of the CREB-UPC. This work integrates in this teleoperation system a set of improvements at a high-level control oriented to improve dexterity, safety and agility of the setup. This improvements are all implemented using a hierarchical based multi-task control. The control improvements are based on redundant and collaborative robots. This work studies the way to exploit the robot kinematic redundancy to increase the dexterity and reachability of the current teleoperation platform and optimize its space occupancy. Moreover, this work also studies collaborative strategies in the surgical workspace, to increase safety in human-robot interactions and optimize the platform preoperative set-up. The outcomes of this work will be integrated in the actual robotic teleoperated platform by replacing the current slave robot for a redundant and collaborative robot
http://hdl.handle.net/2117/333560
ETSEIB-240.152704
Àrees temàtiques de la UPC::Enginyeria mecànica
Integration of a redundant and collaborative robot in a teleoperated fetal surgical platform
oai:upcommons.upc.edu:2099.1/237032021-08-08T01:47:08Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Acevedo Valle, Juan Manuel
author
2014-06
The present work proposes to develop a Fault Tolerant Control strategy based on Robust
Efficient Predictive Control for non-linear systems that can be modeled using a polytopic
model.
The proposed Efficient Predictive Fault Tolerant Control approach relies on the closed-loop Efficient Predictive Control approach that allows to guarantee stability by means of set-invariance and a multilevel hierarchical control approach that guarantees fault tolerance. The lower level
control loop is designed by means of the standard optimal polytopic methodologies based on Linear Matrix Inequalities, enhanced by predictive control layer based on the closed-loop paradigm through invariant-set formulation. In the upper level, fault tolerance is achieved by
means of establishing beforehand a set of bounded faults, against which the proposed Fault Tolerant Control scheme has to guarantee stability and performance. To achieve this, faults are considered as scheduling or uncertain parameters that can vary within a prespecified interval
that corresponds to acceptable fault sizes.
In order to test the proposed approach, the well-known four-tank system benchmark is considered.
The application of the proposed approach implies obtaining a polytopic representation
of the four-tank system that includes the fault effect as extra varying parameters. Different fault scenarios are used to illustrate the performance of the proposed strategy.
http://hdl.handle.net/2099.1/23703
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Energies::Tecnologia energètica::Emmagatzematge i transport de l’energia
Efficient Predictive Control for polytopic systems: application to the four-tank system
oai:upcommons.upc.edu:2099.1/153492020-02-12T20:43:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Cruz Cruzado, Luis Yordano
author
2011-03
El presente tema de tesis presenta un nuevo método de muestreo adaptativo basado en el análisis de componentes principales (PCA). El método de muestreo es iterativo y se basa en ir adaptando periódicamente la base del espacio de configuraciones del sistema robótico usado para la obtención de las muestras, a partir del análisis de componentes principales de las muestras libres de colisión obtenidas hasta el momento. Al método de muestreo se aplicara a un planificador del tipo PRM (Probalistic RoadMap Method).
El desarrollo, se ha realizado en dos partes, la primera parte consiste en implementar el método de muestreo propuesto, realizando ejemplos y pruebas para garantizar la viabilidad de dicho método.
La segunda parte está enfocada en optimizar la planificación de movimientos aplicando un PRM PCA (usa el método de muestreo desarrollado en la primera parte) el cual se aplica a un sistema robótico brazo-mano (brazo robot Stäubli TX90 de seis grados de libertad, equipado con una mano mecánica Shunk SAH de 13 grados de libertad), con una configuración final de aprehensión constreñida.
A sí mismo para realizar una comparación y verificar el rendimiento del PRM PCA, se ha implementado un PRM Gaussiano (basado en muestreo Gaussiano), el cual tiene características similares para solucionar problemas de planificación de movimientos en espacios reducidos o pasajes estrechos. Ambos PRMs (PCA y Gaussiano) han sido bien implementados dentro del entorno de programación para aplicaciones de planificación The Kautham Project, desarrollado en el Instituto de Organización y Control de Sistemas Industriales (IOC-UPC).
http://hdl.handle.net/2099.1/15349
Àrees temàtiques de la UPC::Enginyeria mecànica::Fabricació::Disseny i accionaments de robots
Àrees temàtiques de la UPC::Informàtica::Robòtica
Método de muestreo adaptativo basado en el análisis de componentes principales para la planificación de movimientos de un sistema robótico (brazo-mano)
oai:upcommons.upc.edu:2117/3337242022-06-01T00:27:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Vegas Romero, David
author
2020-09-08
http://hdl.handle.net/2117/333724
ETSEIB-240.143556
Àrees temàtiques de la UPC::Informàtica
Implementation of a Chess Playing robot application
oai:upcommons.upc.edu:2099.1/162972020-02-12T20:43:38Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Muñoz Aguilar, Raúl Santiago
author
2010-01
Two problems related with hybrid electric vehicles have been analyzed in this dissertation.
The first one consists in proposing a propulsion system scheme for the vehicle and the
second one consist in modeling it.
In order to set a propulsion system scheme, the standard configurations for the hybrid
electric vehicles are presented as well as some variations of the series topologies. Then,
a novel configuration which is composed by a synchronous machine and an induction
machine is also presented.
As said before, the second problem consist in modeling this novel configuration.
The reduced model of each machine of the dynamical model is described using the dqtransformation.
Bond Graph and Port-Controlled Hamiltonian approaches are also used
to describe the WRSM, the DFIM and the whole system models.
Finally, the analysis of the power flowing through the system and the regenerative
braking possibilities are done.
http://hdl.handle.net/2099.1/16297
Àrees temàtiques de la UPC::Enginyeria mecànica::Disseny i construcció de vehicles::Automòbils
Àrees temàtiques de la UPC::Enginyeria mecànica::Motors::Motors elèctrics
Modeling and simulation of a series hybrid electric vehicle propulsion system
oai:upcommons.upc.edu:2099.1/153482020-02-12T20:43:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Rodríguez Rosales, Juan Esteban
author
2011-09
This master thesis is aimed at developing a ROS compatible obstacle avoidance module
for a Segway RMP400 platform using lidar sensors. The algorithm implemented supports
different 3D scenarios, such as slopes and precipices. Given that the currently available
navigation modules in the ROS stack work only for 2D, significant modification of the
navigation stack was mandatory. The proposed solution does not take into account the
non-holonomic constraints of the platform. An alternative method is developed to tackle
this issue for the case of tele operated motion.
http://hdl.handle.net/2099.1/15348
Àrees temàtiques de la UPC::Informàtica::Robòtica
Àrees temàtiques de la UPC::Enginyeria mecànica::Fabricació::Disseny i accionaments de robots
Obstacle avoidance for mobile robots using lidar sensors
oai:upcommons.upc.edu:2117/1204692020-02-12T20:43:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Herreras Garcia, Alvaro
author
2018-06-21
In this master thesis, the use of the Takagi-Sugeno and MPC techniques will be applied to the trajectory planning and control of an autonomous vehicle. The first step will be to formulate the model of the autonomous vehicle using a Takagi-Sugeno model. Then, MPC techniques will be used to design a trajectory planner and a kinematic/dynamic controller. The proposed schemes will be tested in a vehicle with the aim of allowing the autonomous driving.
http://hdl.handle.net/2117/120469
ETSEIB-240.134176
Àrees temàtiques de la UPC::Enginyeria elèctrica
Planning and control of an autonomous vehicle using Takagi-Sugeno and MPC techniques.
oai:upcommons.upc.edu:2117/1232092021-09-05T03:12:04Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Martí Saumell, Josep
author
2018-09-17
http://hdl.handle.net/2117/123209
ETSEIB-240.134537
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Hierarchical task control through quadratic programming
oai:upcommons.upc.edu:2117/1083902020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Yang, Ruicong
author
2017-09-13
In this master thesis, the problem of diagnosis an fault-tolerant control of wind turbines will be addressed using Takagi-Sugeno models. Takagi-Sugeno models allow to consider non-linear systems in a linear-like manner by defining a family of models that are scheduled with respect to the operating condition. In the case of the wind turbines, the operating condition is established by the wind.
http://hdl.handle.net/2117/108390
ETSEIB-240.119905
Àrees temàtiques de la UPC::Energies::Energia eòlica
Diagnosis and fault-tolerant control of wind turbines using the Takagi-Sugeno approach
oai:upcommons.upc.edu:2117/1203862022-10-30T02:04:45Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Trapiello Fernández, Carlos
author
2018-06-21
Se explorara la aplicación de técnicas de LPV para el control de un UAV. El controlador será implementado tanto en simulación como en un equipo real. Según las necesidades encontradas se probaran diferentes arquitecturas de control.
http://hdl.handle.net/2117/120386
ETSEIB-240.134079
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Control of an UAV using LPV techniques
oai:upcommons.upc.edu:2117/988942020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ruiz Paz, Diego
author
2016-06-08
In this thesis, a Takagi-Sugeno model for an Attitude/Altitude model of a quadrotor system is developed. With this Takagi-Sugeno model, a gain-scheduling state-feedback controller, as well as a state observer, have been designed for altitude and orientation control. Then, two different control schemes are analyzed in order to have a good performance on tracking changing references in the Attitude/Altitude control. Then an Integral Backstepping controller has been designed for the control of horizontal position. The stability of the whole control system has been studied, and finally the models and controllers have been tested in a simulation environment.
http://hdl.handle.net/2117/98894
ETSEIB-240.120077
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Gain-scheduling Control of a Quadrotor using the Takagi-Sugeno Approach
oai:upcommons.upc.edu:2099.1/151952023-12-03T03:14:51Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Salazar Cortés, Jean Carlo
author
2011-09
El objetivo del presente proyecto final de máster es el diseño e implementación de un sensor de fuerza y par, capaz de medir los tres componentes de fuerza y los tres componentes de par, con alta velocidad de procesamiento de señales que posibiliten su implementación en tiempo real con conexión a la interfaz de usuario a través del protocolo USB, que posea la dimensión apropiada para poder ser montado entre la ultima articulación y el efector final de un robot manipulador. Para lo cual se ha optado por una estructura en forma delta, con un anillo circular exterior y tres vigas radiales espaciadas un angulo de 120º sujetas en el interior a una brida también circular. La estructura se ha mecanizado a partir de un bloque de aluminio, buscando que no existan uniones soldadas o sujetas con tornillos. En cada una de las vigas hay pegadas cuatro galgas extensiométricas que miden la flexión en dos direcciones ortogonales. Son un total de 12 galgas extensiométricas fijadas a la estructura, una por cada cara, que se conectan a seis puentes de Wheatstone para transformar las variaciones resistivas en señales de voltaje. El procesamiento de estas señales, se realiza utilizando un microcontrolador de 16-bits de la familia dsPIC30F de Microchip, luego de realizar el procedimiento de calibración, se pueden obtener las componentes de las fuerzas y pares que actúan sobre el sensor, esta información se transmite a la interfaz de usuario por medio del protocolo USB, que permite y facilita la integración a aplicaciones en tiempo real. Para la validación final del diseño del sensor de fuerza se ha realizado una aplicación de control por fuerza de un robot manipulador SCARA. Con esta aplicación se evalúa el correcto funcionamiento del sensor en aspectos como su velocidad de respuesta y linealidad.
http://hdl.handle.net/2099.1/15195
Àrees temàtiques de la UPC::Enginyeria electrònica::Instrumentació i mesura::Sensors i actuadors
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal
Àrees temàtiques de la UPC::Enginyeria mecànica::Fabricació::Disseny i accionaments de robots
Diseño e implementación de un sensor de fuerza de 6 grados de libertad
oai:upcommons.upc.edu:2117/1092222020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Arbones Clua, Marc
author
2017-09-13
The complexity in modeling human movement increases as the dimensionality of these movement grows.
Since searching more precision and flexibility involves more variables in the model. Dynamic Movement
Primitives (DMP) have shown the ability to generate joint movements with high complexity. However, the
problem remains in the interaction between several joints since DMP alone is not able to deal with it. To solve
this problem a new model called autoencoded dynamic movement primitive (AE- DMP) is introduced in the
work "Efficient movement representation by embedding DynamicMovement Primitives in Deep Autoencoders"[2].
The proposed approach uses autoencoder in order to find a representation of the movement in a latent space.
Consequently, the DMPmodel is able to reconstruct the complete movement. In thisMaster Thesis we will study
the implementation of this model and study its performance. All the features stated in the original paper are
checked, as multiple movements, sparsity and reconstruction of missing or corrupted data.
http://hdl.handle.net/2117/109222
ETSEIB-240.126949
Àrees temàtiques de la UPC::Informàtica
Deep learning: creating bridges between DMPs in autoencoders and recurrent neural networks
oai:upcommons.upc.edu:2117/3749682022-10-25T14:20:18Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Moliner Adell, Eva
author
2022-10-21
Electrolyzers are systems which mission is to produce hydrogen by means of a chemical process called electrolysis. This process is capable of separating the hydrogen and oxygen molecules of the water by applying an electric current I. In that way, the hydrogen is produced in a sustainable way, since it is produced without emitting carbon dioxide into the atmosphere. In order to produce this hydrogen, there are several specifications that have to be accomplished by this high pressure alkaline electrolyzer, such as the difference between the liquid levels in both separation chambers (SCs), or the difference between the pressure of the H2 SC and the pressure of the storage tanks. The aim of this project is to develop a control strategy that achieves the desired specifications in order to increase the supplied gases purity. Since there are multiple objectives which need to be optimized in this problem, a model predictive controller (MPC) based on a multi-objective optimization problem is implemented during the development of this project. To address this problem, it is necessary to work over a control-oriented model of the electrolyzer, as well as over simulation-oriented models. These models are obtained based on several works found in the literature, with high-fidelity models previously implemented and evaluated. During the development of this project, different versions of MPCs are designed and simulated, depending on the tuning and prediction capabilities when disturbances in cur- rent and pressure are considered. The results of these MPCs are compared, as well as the differences yielded by the modification of some parameters of these MPCs. Several key performance indices (KPIs) are considered in order to compare the different versions of MPCs implemented, to facilitate the selection of the one with the best results and also in order to show the effectiveness of the proposed control approach.
http://hdl.handle.net/2117/374968
ETSEIB-240.173239
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Design of predictive controllers for alkaline electrolyzers
oai:upcommons.upc.edu:2117/3326762021-12-16T11:45:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
San Martin Igarza, Ane
author
2020-07-28
Autism Spectrum Disorder (ASD) are neuro-developed mental conditions which mainly affect the interaction in the social area of the affected people. In order
to improve this disorder, different therapies have been developed by therapists, but there is not anyone which could work for all the cases due to the high variety of symptoms that these people can show. However, several studies indicate that they show preference toward interaction with objects, such as computers and robots, rather than with humans. The aim of this work is to develop interactions of a social robot, in terms of actions, movements, sounds and lights in order to interact with children with ASD. These interactions should create a pleasant environment in order to make the children involved in the therapy so that best results can be obtained. In most robots developed until now, two different issues can be observed: the robot is very expensive or the robot is made just to develop an improve a specific skill of the children. Taking these issues into account, our aim is to build a new, fully equipped social robot that has a relatively low cost and is able to interact with children in order to improve different skills of them. We named this robot Lagun. Lagun is equipped with a module v2 camera, speakers, ultrasound sensors, sensitive sensors, LEDs, Arduino Mega and Raspberry Pi 4 which enable the robot to identify the user and its intention so that an adequate action at each time instant can be performed by the robot. The main focus of this work is in the interaction between the child and the robot, which leads the work to an accurate development of the movements, taking care at any moment that the movement being performed is appealing for the child to make this one involved in the interaction. A way to complement this movement, will be by the use of the correct sounds and lights which lead to easier understanding of the state of the robot.
http://hdl.handle.net/2117/332676
ETSEIB-240.152332
Àrees temàtiques de la UPC::Ciències de la salut
An Approach to Attract and Manitain Attention of People with Interactive Mobile Robots
oai:upcommons.upc.edu:2117/1092212019-01-24T10:25:56Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Rodoreda Perales, Carlos
author
2017-09-13
Nowadays, a remarkable development in the automation field is happenning worldwide,
in order to allow to have the most heavy works done by machines. In this topic, it is
starting to be seen the implementation of dual arm robots, able to perform the translation
of heavier and bigger pieces in a most ergonomic way and with better precision, even
giving in many situations the capability to easily pick the most heavy components.
While there is a wide variety of planning tools with a simple and effective interface for
pick and place actions with one arm, the dual arm manipulation is still being seen, currently,
as a very specific and not normalized topic of development, where great part of the
existing tools requires a huge development in order to be apply them effectively, needing
a waste of useless resources (time and capital) to adapt these tools to a new specific robot
or a new environment.
The goal of this project is to create an effective framework in ROS, with a clear and
simple structure (not only for the user, but also for a programmer) which allows to apply
a manipulation, with one or two arms with a wide range of possible configurations.
In order to improve the internal security and usability, this performance will be done enabling
sensors able to understand the environment while a dynamic avoidance of collisions
will be performed.
Great part of this code will be done using the capabilities of the moveit library, in order
to work with a clear and powerful basis, the one will be improved to achieve the already
defined goals.
http://hdl.handle.net/2117/109221
ETSEIB-240.129373
Àrees temàtiques de la UPC::Informàtica::Robòtica
Development and integration of algorithms for manipulation with dual-arm robot systems
oai:upcommons.upc.edu:2099.1/163212023-12-03T03:28:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Paillacho Chiluiza, Dennys Fabian
author
2010-06
Debido a la dificultad de evaluar el rendimiento global de un sistema robótico de control en red de manera analítica, el presente trabajo trata el análisis del comportamiento de un sistema robótico de control en red, tanto temporal como espacial, basado en simulación.
El entorno experimental de sistema robótico de control en red está constituido por un robot manipulador planar con articulaciones angulares, donde el control de movimiento del robot está distribuido en varios controladores remotos, uno para cada articulación, y un controlador principal, los mismos que se comunican por medio de una red de control.
Para simplificar el análisis del modelo del sistema robótico, se ha elegido un modelado por capas, el mismo que permite definir más fácilmente las interacciones entre los componentes del sistema (controlador principal, red de control y controladores remotos).
El modelo del sistema robótico es llevado a un entorno de simulación y puesto a prueba con distintas configuraciones (como el tipo de red de control, tasa de datos, tamaño de datos y número de controladores remotos). De los resultados obtenidos se hace un análisis tanto temporal como espacial del ciclo de comunicación y del error de seguimiento de trayectoria, respectivamente.
http://hdl.handle.net/2099.1/16321
Àrees temàtiques de la UPC::Informàtica::Robòtica
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Análisis del rendimiento de un sistema robótico de control en red basado en simulación
oai:upcommons.upc.edu:2117/3329962024-03-10T06:04:58Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Clemente León, Alejandro
author
2020-07-30
http://hdl.handle.net/2117/332996
ETSEIB-240.152442
Àrees temàtiques de la UPC::Enginyeria elèctrica
Modelling and control of a redox flow batteries
oai:upcommons.upc.edu:2117/806132023-05-19T09:52:48Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Vargas Matín, Elliot
author
2014
This project aims to design and manufacture a bipedal robot speci cally designed to be
controlled with the Inverted Pendulum Foot Placement algorithm. This algorithm, models
the robot as an inverted pendulum. This inverted pendulum is formed from the support
points of the leg of the robot with the ground, to the center of gravity of the robot. Then
using the kinetic and potential energy of the inverted pendulum, the correct position of the
point that represent the center of gravity, is calculated so that the robot does not fall to the
ground.
The project includes the design and manufacture of the robot's electronics, mechanics and
software. But not the implementation of the Inverted Pendulum Foot Placement algorithm
although, a Test Algorithm will be created in order to test that all the mechanics, electronics
and that the robot can be controlled by this type of algorithm.
This robot is powered by lithium batteries and it uses six Dynamixel motors to move. The
input signals are from an IMU (Inertial Measurement Unit) and the encoder inside the motors.
All the electronics are controlled by a microcontroller, but the algorithm is implemented inside
of a Microcomputer. Concerning the mechanics, the entire robot parts were designed in 3D
models and printed using ABS plastic. There is a second design of the robot that is connected
to a PC, as well as, to a power source in order to have in nite autonomy. Also a Stability
Assistance has been made in order to help the algorithm development.
http://hdl.handle.net/2117/80613
Àrees temàtiques de la UPC::Informàtica::Robòtica
Design and manufacture of a biped robot to implement the inverted pendulum foot placement algorithm
oai:upcommons.upc.edu:2117/3717272022-08-24T07:40:13Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Palacín Domínguez, Óscar
author
2022-07-22
The objective of the thesis is to create and implement a generic and modular layer, which encapsulates and coordinates the entire movement of the individual robots that the Mobile Anthropomorphic Dual-Arm Robot (MADAR) platform is made of, to used it in a task and motion planning research framework. As each device has different characteristics and functionalities, is it really important to take them into consideration when a coordinated and synchronized motion planning is wanted. This report tackles two main problems, first, the development of the necessary software for managing the hardware, and second, the synchronization and the control of the set of Real Time (RT) systems. The project uses Robot Operating Systems (ROS), which is a robotic software framework, in a clear and transparent way as a tool to develop the solution for the real hardware motion and communication problems. The framework created enables the user to interact easily with the available functionalities and services that expose the platform. To conclude, using the advantages that the presented solution offers, the project is able to complete a motion planning where the user can have a clear idea of the internal behavior of the involved systems and keep it under control
http://hdl.handle.net/2117/371727
ETSEIB-240.170627
Àrees temàtiques de la UPC::Informàtica::Robòtica
Motion coordination of a dual-arm mobile manipulator
oai:upcommons.upc.edu:2117/3948712023-12-03T03:43:01Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Olvera Marín, Paula
author
2023-09-14
This research focuses on finding solutions to enhance the quality of life for wheelchair users, specifically by applying a gaze-tracking-based interface for the guidance of a robotized wheelchair. For this purpose, the interface was applied in two different approaches for the wheelchair control system. The first one was an assisted control in which the user was continuously involved in controlling the movement of the wheelchair in the environment and the inclination of the different parts of the seat through the user’s gaze and eye blinks obtained with the interface. The second approach was to take the first steps to apply the device to an autonomous wheelchair control in which the wheelchair moves autonomously avoiding collisions towards the position defined by the user. To this end, the basis for obtaining the gaze position relative to the wheelchair and the object detection was developed in this project to be able to calculate in the future the optimal route to which the wheelchair should move. In addition, the integration of a robotic arm in the wheelchair to manipulate different objects was also considered, obtaining in this work the object of interest indicated by the user's gaze within the detected objects so that in the future the robotic arm could select and pick up the object the user wants to manipulate. In addition to the two approaches, an attempt was also made to estimate the user's gaze without the software interface. For this purpose, the gaze is obtained from pupil detection libraries, a calibration and a mathematical model that relates pupil positions to gaze. The results of the implementations have been analysed in this work, including some limitations encountered. Nevertheless, future improvements are proposed, with the aim of increasing the independence of wheelchair users
http://hdl.handle.net/2117/394871
ETSEIB-240.181811
Àrees temàtiques de la UPC::Enginyeria biomèdica::Aparells mèdics
Gaze-tracking-based interface for robotic chair guidance
oai:upcommons.upc.edu:2117/3300612024-03-24T06:08:54Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Civit Bertran, Aniol
author
2020-07-10
The increment of dependant people for the realization of Activities of Daily Living is a fact that comes implicit with the increase of population and their lifespan given by the modern technological progress. For this reason, Assistive Robotics is a field that is currently researched and gradually implemented. In this project, two different algorithms to train robots to perform those tasks are being studied and compared, Reinforcement Learning and Imitation Learning. The main objective is to learn the benefits, drawbacks and suitability of both for a given task. To do the experimentation, the Assistive Gym software will be used, to train the nets with Reinforcement Learning and to perform the movements with Dynamic Movement Primitives learned from a demonstration. The environment will consist in a Sawyer robot cleaning the right arm of a human lying down on a bed.
The potential of each methodology will be tested in different scenarios, having the user’s arm straight or partially flexed. Furthermore, small implementations to improve the execution of the task will be implemented for the Dynamic Movement Primitives to test the performance and the similarity to the demonstration. Both methodologies are capable to perform the desired task successfully, the model trained with Reinforcement Learning is more suitable for an individual user that will use the robot in a long term, due to the training time. The Imitation Learning algorithm is faster to teach and its trajectory and velocities are easily adaptable to the preferences of the user, it is more suitable in places where there is a high rotation of users that require urgent and short attention, such as clinics or hospitals.
http://hdl.handle.net/2117/330061
ETSEIB-240.153759
Àrees temàtiques de la UPC::Enginyeria mecànica
Reinforcement learning for robotic assisted tasks
oai:upcommons.upc.edu:2117/3338552021-04-29T18:57:57Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Pregonas Graells, Jaume
author
2020-07-30
This project is focused on the implementation and analysis of an Electrical Power Steering (EPS) production line. The tool used to achieve the main objective is the ARENA, which is a discreteevent simulation software. Around this finality, there appear different objectives as the implementation of different work methodologies that the production line has. In this document, an analysis of the results obtained from the implementation of a real production line is performed. In addition, it is done a purpose of which implementations could be done in order to improve the production line behaviour. The memory concludes with a brief study of how those improvements could impact economically and environmentally to the production line. Once the implementation and analysis have been done, it can be concluded that the objectives have been achieved. Moreover, it has been possible to improve the production line and increase the production of it.
http://hdl.handle.net/2117/333855
ETSEIB-240.153285
Àrees temàtiques de la UPC::Economia i organització d'empreses::Direcció d'operacions
Study and optimization of a production line for the automotive sector using discrete-event simulation.
oai:upcommons.upc.edu:2117/1748722021-10-10T02:46:28Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Garcia Camacho, Irene
author
2019-10-09
Robotic manipulation of textiles is a very hard task that is receiving lot of attention from the vision and robotics research community. It is hard because perception of textiles should account for many different configurations and varying shapes, and manipulation differs a lot from classical object manipulation in precision and ability. This Master Thesis simplifies the perception part to concentrate in the manipulation problem. Real experiments are performed using a TIAGo robot and custom made grippers. We first realize that manipulation skills require a rather high precision in the positioning, so the hand-eye calibration of the robot is evaluated in the task of grasping a folded cloth. Second, several baseline skills (simple grasps and motions in open loop) are developed and combined to construct two different complex manipulations: spreading a tablecloth and folding a towel. Finally, the manipulations are evaluated using a new proposed benchmark.
http://hdl.handle.net/2117/174872
ETSEIB-240.145249
Àrees temàtiques de la UPC::Informàtica::Robòtica
Robotic manipulation skills for picking and unfolding garments
oai:upcommons.upc.edu:2117/3293332021-12-16T11:45:55Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Martin, Andrea
author
2020-07-22
Service pressure control plays a fundamental role in the framework of water distribution network management, allowing for resource-efficient operation, background leakages reduction and pipe bursts incidence abatement. Surprisingly, however, most water utilities rely on steadystate models for design, supervision and control neglecting the detrimental effects hydraulic transients may have on the system infrastructure. Driven by the industrial case study of the water company Aigües de Barcelona to remedy to the generation of sustained pressure oscillations as a consequence of the operation of the automatic control system, this thesis presents a modelling methodology for transient events based on a finite-difference approximation of the continuity and momentum equations that govern transient flow in pressurised closed conduits. This approach, extended to permit the computation of the pressure losses due to friction accordingly to the Hazen-Williams formula and to account for the presence of branching junctions, of partial pipeline blockages and of dynamically coupled pressure reducing valves, is then applied to model the dynamic behaviour of a section of the water distribution network of the metropolitan city of Barcelona. Furthermore, this thesis proposes a decentralized control scheme constituted by an optimal proportional-integrative controller with anti-windup and a series of model predictive controllers with look-ahead to ensure tracking of the service pressure reference at the entrance of the different district metered areas regardless of the water demand fluctuations. Eventually, the effectiveness of this solution is validated by means of numerical simulations
http://hdl.handle.net/2117/329333
ETSEIB-240.153118
Àrees temàtiques de la UPC::Desenvolupament humà i sostenible::Política i gestió ambiental::Gestió de l'aigua
Transient control in drinking water systems
oai:upcommons.upc.edu:2117/3884702023-06-09T16:40:18Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Mercadé De Luna, María
author
2023-04-20
This project aimed to design and build a low-cost magnetic levitation system based on attraction, using a single coil. The materials and components were carefully selected to keep the cost of the system as low as possible, making it suitable for educational purposes such as laboratory demonstrations or science classes. The system consists of a base made of aluminium and PLA onto which a single coil has been screwed. The intention of the project is to levitate a neodymium magnet underneath the coil by constructing a levitator that works by attraction, i.e. that compensates the force of gravity that the magnet suffers to keep the magnet floating in the air. A PID control algorithm has been implemented using an Arduino microcontroller to regulate the levitation distance between the magnet and the coil. The design and construction of the system were performed with an emphasis on simplicity and affordability. The use of a single coil and inexpensive materials made it possible to achieve a functional prototype at a low cost. The system was tested and the performance of the PID controller was evaluated. The results showed that the system was able to achieve stable magnetic levitation. However, the tuning of the PID parameters was found to be challenging due to the non-linear behavior of the system and an unexpected behavior of the sensor that was expected to give measurements of the magnet position. The sensitivity of the system to external disturbances and noise was also observed. Therefore, a low-cost attraction-based magnetic levitation system was successfully designed and implemented using a single coil, a hall effect sensor and an Arduino microcontroller with a PID control algorithm. The system provides a valuable educational tool for teaching principles of electromagnetics and control engineering. Future work could focus on further improving the system’s performance through the use of more advanced control strategies, such as a sliding mode or fuzzy-logic control
http://hdl.handle.net/2117/388470
ETSEIB-240.170252
Àrees temàtiques de la UPC::Enginyeria electrònica
Construction and control of a magnetic levitation system
oai:upcommons.upc.edu:2117/1701432019-12-11T08:16:06Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Ferrer Sánchez, Jordi
author
2018
Motion control plays an important role in autonomous navigation. It can be de ned as the
compute of motion control inputs for a collision-free path tracking.
To perform the motion control in path planning, a local planner is used. A good local
planner should take into consideration the capabilities of the vehicle (kinematic and dynamic
constraints), in order to compute feasible movements for the robot.
This work is focused on the local planning problem for non-holonomic vehicles in an (x; y; )
space. Three local planners using di erent methods to compute the command velocities are
implemented in the Robot Operating System (ROS) framework, widely used in the robotics
community.
To ful ll our goal, three ROS nodes are created, the rst one uses the Dynamic Windows
Approach algorithm to compute the command velocities. The second uses the Timed Elastic
Band algorithm. The third uses the Model Predictive Control algorithm.
To deal with the non-holonomic vehicle constraints, an Ackermann kinematic model is
used.
IX
http://hdl.handle.net/2117/170143
Àrees temàtiques de la UPC::Informàtica::Robòtica
Implementation and comparison in local planners for Ackermann vehicles
oai:upcommons.upc.edu:2117/3972522023-12-24T04:11:34Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Francés Vigatà, Adrià
author
2023-10-06
En el context de les instal·lacions robotitzades d'última generació, aquest projecte s'enfoca al desenvolupament d'un sistema automatitzat de posicionament d'inserts. L’objectiu principal és millorar l’eficiència de l’estació de clipatge de l’empresa ABB. Per aconseguir això, s'ha dissenyat un sistema que utilitza un robot ABB i una sèrie de dispositius, incloent- hi càmeres, safates vibratòries i tolves. El sistema es basa en la comunicació entre aquests dispositius, coordinant les seves accions per assolir la tasca de clipatge de manera eficient. La càmera té un paper fonamental en proporcionar informació sobre la disposició dels inserts a la safata vibratòria. Aquesta informació s'utilitza per guiar el robot a la selecció i col·locació precisa dels inserts en la seva posició final. El projecte s'ha desenvolupat amb un enfocament a la programació del robot ABB, utilitzant el llenguatge RAPID. S'han implementat rutines i seqüències de moviment perquè el robot pugui recollir i col·locar els inserts de manera eficient. A més, s'ha creat una interfície de control que permet ajustar els paràmetres del sistema, com ara les seqüències de vibració de la safata vibratòria i els òfsets de recollida dels inserits. L'estructura de connexions de la controladora OmniCore d'ABB ha estat clau en el funcionament del sistema. S'ha establert una distinció entre xarxes públiques i privades per garantir un control adequat de qui té accés a les dades i senyals dels diferents dispositius. Al llarg del projecte, s'han superat diversos desafiaments tècnics, com ara el calibratge de la càmera i l'optimització de les seqüències de vibració de la safata. S'ha treballat en la creació d'una interfície d'usuari intuïtiva que facilita la configuració i l'ajust de paràmetres. En conclusió, aquest projecte ha desenvolupat un sistema automatitzat de posicionament d'inserits que millora significativament l'eficiència de l'estació de clipatge d'ABB. S'ha aplicat una programació efectiva al robot, s'ha establert una estructura de connexions sòlida i s'ha creat una interfície de control versàtil. El projecte representa un aprenentatge valuós en el camp de la robòtica i l'automatització industrial, i estableix les bases per a futures millores i desenvolupaments en aquest camp
En el contexto de las instalaciones robotizadas de última generación, este proyecto se enfoca en el desarrollo de un sistema automatizado de posicionamiento de insertos. El objetivo principal es mejorar la eficiencia de una estación de clipaje de la empresa ABB. Para lograr esto, se ha diseñado un sistema que utiliza un robot ABB y una serie de dispositivos, incluyendo cámaras, bandejas vibratorias y hoppers. El sistema se basa en la comunicación entre estos dispositivos, coordinando sus acciones para lograr la tarea de clipaje de manera eficiente. La cámara desempeña un papel fundamental al proporcionar información sobre la disposición de los insertos en la bandeja vibratoria. Esta información se utiliza para guiar al robot en la selección precisa de los insertos. El proyecto se ha desarrollado con un enfoque en la programación del robot ABB, utilizando el lenguaje RAPID. Se han implementado rutinas y secuencias de movimiento para que el robot pueda recoger y colocar los insertos de manera eficiente. Además, se ha creado una interfaz de control que permite ajustar los parámetros del sistema, como las secuencias de vibración de la bandeja vibratoria y los offsets de recogida de los insertos. La estructura de conexiones de la controladora OmniCore de ABB ha sido clave en el funcionamiento del sistema. Se ha establecido una distinción entre redes pública y privada para garantizar un control adecuado de quién tiene acceso a los datos y señales de los diferentes dispositivos. A lo largo del proyecto, se han superado varios desafíos técnicos, como la calibración de la cámara y la optimización de las secuencias de vibración de la bandeja. Se ha trabajado en la creación de una interfaz de usuario intuitiva que facilita la configuración y ajuste de parámetros. En conclusión, este proyecto ha logrado desarrollar un sistema automatizado de posicionamiento de insertos que mejora significativamente la eficiencia de la estación de clipaje de ABB. Se ha aplicado una programación efectiva en el robot, se ha establecido una estructura de conexiones sólida y se ha creado una interfaz de control versátil. El proyecto representa un aprendizaje valioso en el campo de la robótica y la automatización industrial, y sienta las bases para futuras mejoras y desarrollos en este campo
In the context of state-of-the-art robotic installations, this project focuses on the development of an automated insert positioning system. The main objective is to improve the efficiency of a clipping station of the ABB company. To achieve this, a system has been designed using an ABB robot and a series of devices, including cameras, vibrating trays and hoppers. The system relies on communication between these devices, coordinating their actions to achieve the clipping task efficiently. The camera plays a key role in providing information about the arrangement of the inserts in the vibratory tray. This information is used to guide the robot in the precise selection of the inserts. The project has been developed with a focus on programming the ABB robot, using the RAPID language. Routines and motion sequences have been implemented to enable the robot to pick and place the inserts efficiently. In addition, a control interface has been created to allow adjustment of system parameters such as vibratory tray vibration sequences and insert pick-up offsets. The connection structure of ABB's OmniCore controller has been key to the operation of the system. A distinction has been made between public and private networks to ensure proper control of who has access to data and signals from the different devices. Throughout the project, several technical challenges have been overcome, such as camera calibration and optimization of the tray vibration sequences. Work has been done to create an intuitive user interface that facilitates the configuration and adjustment of parameters. In conclusion, this project has succeeded in developing an automated insert positioning system that significantly improves the efficiency of ABB's clipping station. Effective programming has been applied to the robot, a robust connection structure has been established, and a versatile control interface has been created. The project represents valuable learning in the field of robotics and industrial automation, and lays the foundation for future improvements and developments in this field
http://hdl.handle.net/2117/397252
ETSEIB-240.181153
Àrees temàtiques de la UPC::Informàtica::Robòtica
Automated insert positioning system to increase efficiency and accuracy in an ABB clipping station
oai:upcommons.upc.edu:2117/1329692019-06-20T08:07:07Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Desquins, Théo
author
2018
http://hdl.handle.net/2117/132969
Àrees temàtiques de la UPC::Informàtica::Robòtica
Construction de modèle par similitude et reconnaissance de pièces
oai:upcommons.upc.edu:2117/3546372022-05-17T17:00:21Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Miranda Moreno, Antonio
author
2021-10-11
An a priori map is often unavailable for a mobile robot in an unknown environment. In large spaces, relying on a tele-operated guidance to create a map of the environment will result in a huge and inefficient workload. Thus, an autonomous exploration algorithm is required for a mobile robot to complete the exploration. The work proposes an autonomous exploration and mapping method for a tracked-robot based on the information gained from a LiDAR sensor, the motion cost, and a grid map that represents the built map for navigation and localization. The project is created from scratch, that means that some previous work has to be done before implementing and testing the algorithm on the robot. The previous work carries out a simultaneous localization and mapping while searching for interesting objects that may modify the navigation of the robot, the navigation planner that generates trajectories for the robot and the kinematic model of the tracked-robot. Once all those steps are implemented, the experiments can be done on the real robot, and the exploration strategy is tested both in the real robot and in the simulator. All the project has been implemented using ROS. Section 2 gives a brief introduction of the history of tracked robots, SLAM and autonomous exploration. Then, Section 3 covers the SLAM approach used in the project, how to parametrize it in ROS and the detection and localization of stairs and holes in the map. After that, Section 4 introduces the autonomous exploration strategy used and the algorithms used to compute the set of goals required for an efficient and optimized exploration. Section 5 explains how the global and local planners are implemented and the costmap layers. Hardware and Software required for making everything work is presented in Section 6 and 7 respectively. Here the tracked robot and the sensors used are presented, as well as the kinematic and control models for the robot and the simulator used for the different maps tested. Finally, Sections 8, 10 and 10 explains the different problems the project had to deal with, and the experiments run in the real robot and the simulator. Once the results are presented, conclusions are commented.
http://hdl.handle.net/2117/354637
ETSEIB-240.160943
Àrees temàtiques de la UPC::Enginyeria electrònica
Exploration and search in unknown open environments with a mobile robot
oai:upcommons.upc.edu:2117/1273532023-12-03T03:57:11Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Galve Ceamanos, Sergio
author
2018-09-17
Everyday robots are more involved in our daily basis and they are expected to be part of the domestic environment eventually. Assertive robots deal with this scenario providing help to people with certain disabilities, a task that requires an intuitive communication between human and robot. One of the control methods that have become popular is the Brain Computer Interfaces (BCI), using electroencephalography (EEG) signals to read the user’s intention. Commonly applied to let the user choose among several options, without further interaction once the robot starts acting. This project explain a method to interpret the EEG signal online and use it to manipulate online the movement of a robot arm. The signal that is received comes from Motion Imagery to allow the user to communicate constantly his intention. Using Dynamic Movement Primitives (DMP) and Virtual Force Systems the stored trajectories of the robot can be modified online trying to adapt to the user’s will. With this elements the trajectories are applied in a real robot arm, chequing online if all the requested goals are feasible positions for the robot. This method intend to make more natural the collaboration with a robot in domestic tasks, where slight modifications of an action can lead into a more satisfactory interaction.
http://hdl.handle.net/2117/127353
ETSEIB-240.134840
Àrees temàtiques de la UPC::Informàtica::Robòtica
EEG based volitional interaction with a robot to dynamically replan trajectories
oai:upcommons.upc.edu:2117/1768422020-03-03T16:53:29Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Jouin, Thomas
author
2019-10-09
This thesis fell within the project by the Borobo company located in Nice, France, to build an autonomous mobile robot for load transportation for elderly people and people with limited mobility at a limited cost. The objective of this thesis was to implement Simultaneous Localization and Mapping (SLAM) and autonomous navigation on a four-wheeled mobile robot using low cost sensors. The method used to do so was first to list some sensors with limited cost used as odometry source and obstacle detection devices. The implementation of those sensors on the robot was also explained. Then SLAM algorithm and autonomous navigation were implemented on the robot using the Robot Operating System (ROS) and some packages provided by the ROS community. Finally the performances of SLAM and autonomous navigation were assessed for different configurations of sensors and parameters. The advantages and disadvantages of each sensor and navigation methods were listed. Regarding the odometry sources, wheel odometry from rotary encoder, visual odometry from Intel RealSense Depth camera D430 and odometry from Intel RealSense Tracking camera T265 were compared. The wheel odometry of this four-wheeled differential drive robot showed significant error due to the slippage of the wheels on the ground when executing a rotational movement of the robot. This error was limited for the visual odometry from the D430 camera but the fragility of this sensor made it unusable. The most precise odometry source tested was the T265 tracking camera which in addition to providing a precise odometry thanks to its embedded Inertial Measurement Unit (IMU) and visual-SLAM processing, was relieving the main processing unit of the robot from visual computing. For obstacle detection, the RPLidar A1 lidar sensor and the Intel RealSense Depth camera D430 were assessed. On the one hand, the lidar sensor detected the obstacle in only two dimensions but with a great precision and with a 120°field of view. On the other hand, the depth camera could detect obstacles in three dimensions but with great imprecision and limited field of view of only 85.2°. The SLAM algorithm executed with the lidar was precise and showed good close loop detection whereas with the depth camera, the depth data were too noisy to be exploitable. After completing these comparisons it was stated that the best configuration of sensors for the prototype of the Borobo robotwas to use the Intel RealSense Tracking camera T265 as odometry source and the RPLidar A1 lidar as a laser scan source. However, further developmentwas required to be completed on wheel odometry using an IMU to correct the rotational error. Further tests were also advised on the Intel RealSense depth camera D435, a less fragile version of the D430 camera, for odometry source but especially for obstacle detection as it allow a low cost obstacle detection in three dimensions. Finally some improvements and parameter tuning were recommended on the navigation algorithm.
http://hdl.handle.net/2117/176842
ETSEIB-240.146004
Àrees temàtiques de la UPC::Informàtica::Robòtica
Implementing SLAM and navigation on an autonomous mobile robot for load transportation for disabled and elderly people.
oai:upcommons.upc.edu:2117/998152020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Gómez Bella, Guillermo
author
2016-01
http://hdl.handle.net/2117/99815
Àrees temàtiques de la UPC::Energies
Automatic method for the reduction of load cases in structural fatigue calculations applied to wind turbine components
oai:upcommons.upc.edu:2117/3710762022-07-26T08:00:28Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Fontal Chacón, Adrián
author
2022-06-20
Modern factories are incorporating robotics and automatics in their processes. From robotic arms to Automated Guided Vehicles (AGVs), the goal of these machines is helping humans to do the hardest or more repetitive tasks, increasing productivity. One of the most used robotic machines are AGVs. These robots help factories like car factories to automate their processes, for example moving pieces from the warehouse to the assembly line when necessary. This project will review a computer vision method called Visual Servoing applied in a Turtlebot 2 platform, with the objective to adapt it to the docking of an AGV. This will be done using a combination of the two ways to perform visual servoing, Position Based Visual Servoing (PBVS) and Image Based Visual Servoing (IBVS), along with different computer vision techniques such as binarization, equalization or noise removal, and different aspects in robotics analysis such as study of its kinematics, or transformations between coordinate frames, which will be essential tools in order to do this project. Once the project is done an analysis of future work and what has to be improved about this solution will be done, along with a study of its budget and environmental impact.
http://hdl.handle.net/2117/371076
ETSEIB-240.172390
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
Visual-servo trajectory optimization for AGVs
oai:upcommons.upc.edu:2117/1286212020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Fuerfanger, Bernhard
author
2019-01-17
This Master thesis discusses how to obtaining a accurate nonlinear model of a cus-tomized quadcopter with gimbal attachment in order to control such with a nonlinearmodel predictive controller. This approach uses a state dependent coefficient form tocapture the UAV’s nonlinearities obtained from a series of analyses and validated withthe motion capture system OptiTrackR©. To obtain the model, data is collected in testflights with manual control and used as a input for the system identification toolbox inMATLABR©. The nonlinear model predictive control law is derived by using a sequentialquadratic programming solver (CMSCGMRES) while accounting for input, output andstate constraints. The performance is shown through several representative flights withaggressive reference changes and minimum disturbances
http://hdl.handle.net/2117/128621
ETSEIB-240.136143
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Modelling and control of a customized drone with gimbal-attachment
oai:upcommons.upc.edu:2117/3508632022-04-13T05:24:56Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Legarda Herranz, Guillermo
author
2021-06-21
Swarm robotics, inspired by natural swarms, studies how simple robots with only local sensing capabilities and no centralised control may cooperate to achieve a common goal in a robust, flexible and scalable way. A robotic system with such properties constitutes an interesting alternative to the platforms currently used in warehouses and distribution plants, where workers are at risk of injury and the space and budget available for complex infrastructure is limited. Swarm behaviours are emergent, which makes the task of designing the controllers of the individual robots particularly challenging. In this work, we propose a method for a swarm of industrial robots to collectively transport items that are too heavy for a single agent to carry. We use artificial evolution to evolve behaviour tree controllers for the swarm agents and we conceive a decentralised coordination strategy based on local messaging. The method is developed and tested in a simulated environment, using a combination of freely available open source libraries. The results show that a homogeneous swarm equipped with our solution is able to successfully find the items placed in the environment and transport them back to a nest region. We suggest further tuning of the evolutionary parameters and the introduction of noise in the simulator in order to improve the observed performance of the controllers in simulation and their expected performance the real world
http://hdl.handle.net/2117/350863
ETSEIB-240.160390
Àrees temàtiques de la UPC::Enginyeria electrònica
Evolution of behaviour trees for collective transport with robot swarms
oai:upcommons.upc.edu:2117/3311712021-12-16T11:45:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Amenabar Echave, Leire
author
2020-07-10
In an ever-changing and demanding world new technologies, which allow more efficient and easier industrial processes, are needed. Furthermore, until now, traditional vision technologies and algorithms have been used in the industrial area. These techniques, even though they achieve good results in simple vision tasks, they are really limited since any change in the processed image affects their performance.
For example, in code reading tasks, if the code has a mark or it is not completely visible, the piece with the code would be discarded which leads to losses for the company. These kind of problems can be solved by the implementation of machine learning techniques for vision purposes. Moreover, these techniques learn from example images and even though a perfect performance is difficult to get, machine learning techniques are much more flexible than traditional techniques. Even though the term machine learning was coined for the first time in 1959, until now, these techniques have barely been implemented in the industrial area. They have mostly been used for investigation purposes.
Apart from the new vision techniques, new types of robots are being implemented in industrial environments such as collaborative or social robots. On the one hand, collaborative robots allow the workers to work next to or with the robot without any type of physical interference between them. On the other hand, social robots allow an easier communication between the robot and the user which can be applied in different parts of the industry such as introducing the company to new visitors. The present project gathers information in regard to the analysis, training and implementation of a vision artificial neuronal network based software called ViDi Cognex software. By the use of this software, three different vision tasks were trained. The most important one is the hand gesture recognition task since the obtained hand gesture controls the action performed by the YuMi robot, which is programmed in RAPID language. It is believed that the development of the different artificial neuronal networks with industrial purposes can show the applicability of machine learning techniques in an industrial environment. Apart from that, the hand gesture recognition shows an easy way to control the movements of a robot which could be used by a person with no knowledge of robots or programming. To finish, the use of a two arm collaborative robot, could show the potential and versatility of collaborative robots for industrial purposes.
http://hdl.handle.net/2117/331171
ETSEIB-240.154252
Àrees temàtiques de la UPC::Enginyeria mecànica
Real-Time Collaborative Robot Control Using Hand Gestures Recognised By Deep Learning
oai:upcommons.upc.edu:2117/841722024-01-14T04:58:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Repiso Polo, Ely
author
2015
This Master’s Thesis presents a developed tracking algorithm which fuses different object
detections. These object detections are obtained from different detectors. These detectors are: a
laser leg detector and a tag vision detector. The fusion has the objective to obtain more robust
object detections; moreover, it has the capability to distinguish specific people. In the future,
it is intend to add new detectors, such as: a skeleton person detector and a radar detector. Adding
new detectors is direct and easy, since the tracker has a general and flexible implementation which
allows the change of the type and number of detectors.
The Robust multi-hypothesis tracker fusing diverse sensor information uses as the back- bone of the
theory the multi-hypothesis tracker approach proposed by Reid [28]. Then, the algorithm has been
modified to add some new improvements.
For work with social robots in the algorithm was added two functions in the probabilities of the
hypothesis which allows us to control the confirmation and the deletion of the tracks. Also, the
algorithm was improved by adding another function that lets we use the people’s velocities
orientations to improve the association between tracks and detections in crossing situations.
Furthermore, the algorithm was extended with the fusion of different detections.
For work with autonomous vehicles in the European Project Cargo ANTs1 the algorithm was adapted for
tracking locally (without map) and global (with a map); was improved to use the track velocity and
a specific detector to distinguish between dynamic and static objects and was adapted to group
different objects of the same type using distance. The objects can have dynamic or static type.
Finally, the tests of the tracker algorithm have been carried out in the fields of social robots
and autonomous vehicles (Cargo ANTs project). The experiments have been carried out with Tibi and
Dabo robots, illustrate that our tracking approach can robustly and efficiently track multiple
people without changes on their id ’s and without losing them. Also, the experiments show that the
tracker can distinguish a specific person by using the fusion; even in situations which has several
people or objects that cross with or occlude the specific person. The experiments have been carried
out in the Cargo ANTs project show that the tracker is able to follow only the single real dynamic
object, and that the tracker is able to group different detections obtained inside the same object.
http://hdl.handle.net/2117/84172
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal
Àrees temàtiques de la UPC::Enginyeria electrònica::Instrumentació i mesura::Sensors i actuadors
Robust multi-hypothesis tracker fusing diverse sensor information
oai:upcommons.upc.edu:2117/775602020-02-12T20:43:39Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Torren Larroya, Abel
author
2015-03
This master thesis addresses the LPV control of a quadrotor system. The quadorotor model
is first transformed to a LPV model respresentation starting from the nonlinear physical model.
Then, using the LPV gain-scheduling control theory, several LPV controller schemes are designed.
They are based on static state feedback control with or without reference model. Stability and
performance will be established by means of LMIs. Finally, the implemented control solution will
be tested in a quadorotor system in a simulation environment using the non-linear model. To
assess the performances of those controllers, several scenarios have been simulated.
http://hdl.handle.net/2117/77560
Àrees temàtiques de la UPC::Aeronàutica i espai
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
LPV Control of a Quadrotor
oai:upcommons.upc.edu:2117/1674872019-09-23T10:06:06Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Ferrandis Vilaró, Joan
author
2018
http://hdl.handle.net/2117/167487
Evaluate and correct robot actions using EEG signals
oai:upcommons.upc.edu:2117/3328552023-06-30T00:27:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Daïni, Xavier
author
2020-09-14
In the modern world, we have many ways to localize robots in space. The most known methods are based on GPS or triangulation, and the orientation sensors also give good results (Angle of Arrival : [Cin+ts] GPS/Bluetooth trilateration : [SS18]). However these methods, although helping in most situations, also have their problems. GPS and triangulation have issues when we have not access to the satellites (like an indoor application) and orientation sensors have little deviation overtime. Furthermore the two first doesn’t give an easy access to the heading (orientation) of the device unless another sensor is placed on said device (in which case we have the direction in which the object is and a third is needed to entirely know the orientation). These facts have led some companies and our laboratory to investigate on some new ways to make an indoor position sensor, which would be complementary to the already existing methods in order to reduce their drawbacks. In order to achive this goal, the team has developed a new couple of beacon and sensor, which
would act as a “shower”. When the robot passes near the zone covered by the beacon, then its sensor is activated and the robot actualizes his position, with no errors. It is based on radiometric properties of light in order to estimate the distance between one led of the beacon and one pixel on the sensor. Knowing that the sensor has 3 pixels disposed in an L-shape, and that there are 3 leds on the beacon, each one of them emitting at the same wavelength but at different frequencies of emission (5kHZ, 11kHz and 17kHz) the sensor is able to determine which led it is able to read and so with some mathematical models and an Extended Kalman Filter, we were able to estimate with a good precision, the position of the robot. Along the course of the project, due to the covid-19 crisis, a new field of studies has emerged, the usage of the optical-flow.
http://hdl.handle.net/2117/332855
ETSEIB-240.154139
Àrees temàtiques de la UPC::Enginyeria mecànica
Integration of a bio-inspired optical sensor on a mobile robot for indoor localization
oai:upcommons.upc.edu:2099.1/163172021-11-14T04:51:20Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Gámiz Caro, Javier Francisco
author
2011-09
En la optimización del funcionamiento de los procesos que intervienen en una estación
depuradora de aguas residuales es importante aunar la sinergia entre estrategias de control
automático, diseño ergonómico de centros de control e interfaces hombre-máquina, y cuidado
del medio ambiente.
Existen en el mercado herramientas de modelado y simulación que ofrecen al experto ayudas
para el soporte a la decisión en este ámbito, una interfaz simulada de control y supervisión
puede aportar también aspectos de creación de prototipos y diseños de nuevas estrategias de
control de una forma rápida y eficaz, permitiendo incluir la valoración del experto en la mejora
de los sistemas.
El presente trabajo de máster ha consistido en el diseño de una interfaz simulada que permita
evaluar los diferentes comportamientos de técnicos y responsables del telecontrol de
Estaciones de Tratamiento de Agua. Dicho entorno tecnológico está destinado también a ser
una herramienta de validación de nuevas interfaces que se deseen integrar en programas de
supervisión y control en explotación.
La arquitectura utilizada en el diseño e implementación de la interfaz de simulación está
basada en un software de emulación en la plataforma ControlLogix de Rockwell Automation
para la lógica de control del controlador (PLC), un software de supervisión y control (SCADA)
Intouch de la firma WonderWare y la tecnología OPC/DDE como interfaz estándar de
comunicaciones entre el PLC y el Scada.
http://hdl.handle.net/2099.1/16317
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Desenvolupament humà i sostenible::Enginyeria ambiental::Tractament de l'aigua
Diseño e implementación de una interfaz simulada para el estudio del control y supervisión de una estación depuradora de agua residual
oai:upcommons.upc.edu:2117/979102020-02-12T20:43:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Guerrero Font, Eric
author
2016-09-05
This Master’s Thesis details the practical implantation of an acoustic positioning and communication
system in an autonomous underwater vehicle used for research purposes. The implementation has
been followed by the evaluation of the system performance. Both, in simulation and in field tests. The
purpose of this integration, is to provide an absolute position measurement and a continuous communication
link when the robot is underwater, in order to increase the localization accuracy and the robot
supervision. It will allow to perform larger and deeper missions, increasing the autonomy of the robot.
This chapter presents the project by first exposing the motivation, section 1.1, secondly it enumerates
the goals of the project in the section 1.2, and thirdly the research projects in which Thesis is framed,
section 1.3. Moreover, an state of the art is presented in the section 1.4, followed by the methodology
used, section 1.5.
http://hdl.handle.net/2117/97910
ETSEIB-240.121062
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
USBL integration and experimental assessment in a multisensor EKF-based AUV navigation approach.
oai:upcommons.upc.edu:2117/3328162021-12-31T01:28:06Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Gómez-cambronero Martín, Pablo
author
2020-07-31
Robotics is changing the world in a fast manner, simplifying complex and tedious tasks not only in the industrial processes and military fields but also in a social scope and in other areas as, for instance, the farming field that, until now, have been using rudimentary technologies and techniques for conducting its purposes.
Earth Rover Europe S.L. is a company within the technological agriculture sector which is developing novel solutions with the aim of helping to feed a growing global population by improving yields, reducing on farm crop waste and at the same time reducing CO2 emissions. Among these contributions is the development of unmanned ground vehicles which will work autonomously inside a farm, carrying out a set of activities previously performed by humans in a less efficient and ecological manner.
At this point is where the project presented in the current document takes place. This thesis, which has been conducted inside the previously mentioned company, will present the fundamentals of the development of the central system that will be in charge of the supervision of the entire farm and its elements as well as of the orchestration of the fleet of robots that will be working in it. In order to achieve this, a meticulous description of the entire system, its components and the relations between them must be done before getting on with its implementation. This central system includes a global planning algorithm which optimizes the routes taken by the different robots within the farm to accomplish with the assigned missions. Furthermore, this system must be integrated within an operational environment, therefore, this document will also present the developed simulation for testing purposes and also the real implementation of one rover for doing some real experiments.
http://hdl.handle.net/2117/332816
ETSEIB-240.152426
Àrees temàtiques de la UPC::Informàtica
Development of a mobile robot mission scheduler system based on high level planned farming activities
oai:upcommons.upc.edu:2117/1085542020-02-12T20:43:41Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Alvarado Vargas, Juan Elwin
author
2017-09-13
This project focuses in the initialization of the work and in the study of online services in order to design and implement an automatic speech recognition system for the robotic platform MASHI. This system will be implemented in two Raspberry Pi 3 using a Master-Slave structure. Online resources and services will be used to maintain the wireless connection and control of the platform. As the desired functionality, this automatic speech recognition system will serve as an efficient interface for the interaction between MASHI and the people inside public buildings, the interaction of the system with other interconnected devices is also considered.
http://hdl.handle.net/2117/108554
ETSEIB-240.126118
Àrees temàtiques de la UPC::Informàtica
Design and implementation of an automatic speech recognition interface for a Multipurpose Assistant Robot (MASHI)
oai:upcommons.upc.edu:2117/3323732021-12-16T11:48:03Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Mora Martínez, Alejandro
author
2020-07-16
Autonomous robots are highly complex systems. In order to operate in dynamic environments, adaptability in their decision-making algorithms is a must. Thus, the internal and external information that robots obtain from sensors is critical to re-evaluate their decisions in real time. Accuracy is key in this endeavor, both from the hardware side and the modeling point of view. In order to guarantee the highest performance, sensors need to be correctly calibrated. To this end, some parameters are tuned so that the particular realization of a sensor best matches a generalized mathematical model. This step grows in complexity with the integration of multiple sensors, which is generally a requirement in order to cope with the dynamic nature of real world applications. This project aims to deal with the calibration of an inertial measurement unit, or IMU, and a Light Detection and Ranging device, or LiDAR. An offline batch optimization procedure is proposed to optimally estimate the intrinsic and extrinsic parameters of the model. Then, an online state estimation module that makes use of the aforementioned parameters and the fusion of LiDAR-inertial data for local navigation is proposed. Additionally, it incorporates real time corrections to account for the time-varying nature of the model, essential to deal with exposure
to continued operation and wear and tear. Keywords: sensor fusion, multi-sensor calibration, factor graphs, batch optimization, Gaussian Processes, state estimation, LiDAR-inertial odometry, Error State Kalman Filter, Normal Distributions Transform.
http://hdl.handle.net/2117/332373
ETSEIB-240.152465
Àrees temàtiques de la UPC::Enginyeria mecànica
Robust 3D IMU-LIDAR Calibration and Multi Sensor Probabilistic State Estimation
oai:upcommons.upc.edu:2117/775552020-07-15T04:20:46Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Riba Pi, Edgar
author
2015-06
http://hdl.handle.net/2117/77555
Àrees temàtiques de la UPC::Informàtica
Implementation of a 3D pose estimation algorithm
oai:upcommons.upc.edu:2117/867362020-02-12T20:43:39Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ajenjo Escolano, Enrique Javier
author
2016-01
Until now, the robots environment was characterized for being parametrized. This means, that all tasks
must be defined and the robot only needs to repeat the programed action continually. Furthermore, the
robot and the user don't share the same workspace. But now, new environments have appeared, bringing
a new robotic paradigm where the robots and the users share the same workspace in semistructured
environments. The tasks in this new paradigm have a plasticity, where they evolve depending on the
user. It is therefore important that the system evolves at the same time than the user.
In this project, a solution of an activity recognition system to be used in a robotic assistance on
domestic environments is presented. This system works in real-time to monitor the user's activities
at the same moment that the action is occurring, computing an early prediction and offering to the
system future information about the possible necessities of the user's to finalize the activity that the
system predicted.
The solution proposed in this project must solve the following sub-objectives:
1. Object Segmentation: The system separates the objects from the scene. We used one model
of scene to compare the differences between the model and the actual frame.
2. Object Recognition: The system recognizes what objects are involved in the action. To
capture the characteristic points of the objects, we first use the SURF Key-Point of the objects
and, then, we convert this information in the Bag of Features representation. In the second step
of the process, we use a Support Vector Machine (SVM) to build the classifier structure.
3. Object Tracking: The system obtains the dynamic information of the objects in the scene.
The system uses a CamShift algorithm to monitor the state of the objects.
4. Activity Recognition: The system identifies the action that the user is doing taking the
previous information into account.
Other important issues we need to take into account in the development of this project are:
· Scalability: The system needs to be able to grow so that it can consider more cases. The
installation of new hardware or the use of new algorithms are two examples of this.
· Working in Real Time: The system needs to predict the actions at the same time than they
are happening.
· Consideration of Different Type of Objects: In domestic scenarios, different types of
objects with different characteristics need to be considered. For example, labeled objects, smooth
surfaces with a dominant color and textured objects.
· The Use of Standard Robot Operation System: We use the Indigo ROS version to develop
this project and share the intermediate information of the project so that other projects can use
it.
The results obtained applying all the process is an accuracy detection of the activity only using the
static information. The dynamic information we used to increase the detection of the more complex
activities or more complex situations, and the actual version of the project grows in complexity with
respect to the previous project but it needs to grow more to take the results of this project into account
to increase the trustworthiness of the project.
http://hdl.handle.net/2117/86736
Àrees temàtiques de la UPC::Informàtica
On-line human activity recognition by monitoring the interaction with the objects in a known environment
oai:upcommons.upc.edu:2099.1/153472020-02-12T20:43:37Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Trulls Fortuny, Eduard
author
2009-09
La transición de escenarios controlados en condiciones de laboratorio al mundo
real conlleva nuevos desafíos en el campo de la robótica móvil, en aspectos tan
básicos como la navegación autónoma. En este proyecto se propone un método
de navegación, evitando obstáculos, para plataformas móviles en entornos urbanos
y dinámicos. El trabajo está enmarcado en el contexto de un proyecto de
investigación y debe actuar como un módulo independiente que recibe como entrada
datos de los sensores del robot y metas expresadas en coordenadas locales,
provenientes de un módulo de un nivel de abstracción superior, y produce como
resultado instrucciones a los actuadores de los motores del robot. Se presenta
un método de navegación que utiliza un planificador local para determinar el
camino desde un punto inicial a una meta, y a partir de los puntos encontrados
por el planificador, sigue el camino usando un sensor láser para detectar
los obstáculos en el camino. Si hay obstáculos corrige la trayectoria inicial para
encontrar un nuevo camino sin obstáculos. El controlador de movimiento, que
calcula velocidades de traslación y rotación que satisfagan las restricciones dinámicas
del robot y lo acerquen a la meta siguiendo el principio de la ventana
dinámica. El planificador local no usa métodos de extracción de características
por la carga computacional que requieren, pero sí de filtrado, consiguiendo
tiempos de respuesta muy pequeños, adecuados para el control reactivo de un
robot en tiempo real. Adicionalmente se estudia el problema de la navegación
en entornos urbanos con escaleras, agujeros, rampas y desniveles, mediante el
uso de un sensor láser volteado 90 para obtener un perfil del espacio enfrente
del robot e inferir si es navegable. Nuestra propuesta ha sido implementada en
dos robots Segway.
http://hdl.handle.net/2099.1/15347
Àrees temàtiques de la UPC::Enginyeria mecànica::Fabricació::Disseny i accionaments de robots
Àrees temàtiques de la UPC::Informàtica::Robòtica
Diseño e implementación de un sistema de evasión de obstáculos para robots móviles en entornos urbanos
oai:upcommons.upc.edu:2099.1/163222021-09-05T06:06:31Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Carrero Candelas, Niliana Andreina
author
2010-05
En este trabajo se presentan herramientas interactivas en 2D y 3D para ilustrar los conceptos
teóricos de control en modo de deslizamiento de sistemas dinámicos lineales a trozos. El objetivo
final es permitir a los alumnos la posibilidad de visualizar la evolución del sistema y su
respuesta en tiempo real, ante los cambios y configuraciones de toda una serie de parámetros
del sistema en estudio. De esta manera, podrán obtener conocimientos prácticos y avanzar en la
compresión de la respuesta del sistema de manera cualitativa y cuantitativa. Las herramientas
se han desarrollado mediante Sysqueke y Easy Java Simulations (EJS), que permite una fácil
programación y diseño.
http://hdl.handle.net/2099.1/16322
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Informàtica::Sistemes d'informació::Interacció home-màquina
Àrees temàtiques de la UPC::So, imatge i multimèdia::Dispositius de so, imatge i multimèdia
Herramientas interactivas para control en modo de deslizamiento
oai:upcommons.upc.edu:2117/3513152021-12-16T11:51:13Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Gil Delgado, Carlos
author
2021-06-21
Muscular Dystrophies refers to a group of genetic diseases which cause progressive weakness and degeneration of muscles used during voluntary movement. Many affected people use powered wheelchairs, increasing the muscle loss due to the lack of activation of the upper limb muscles. In order to improve the muscular condition of these subjects was born the powered wheelchair MOVit project. This wheelchair is controlled moving back and forth both arms. The objective of this thesis is to develop an entertaining interface composed by a proof-of-concept car robot, named MOVit Lite, for a future application in the MOVit system controlled with the GripAble hand-training tools. The rehabilitation interface, together with different control modes, have been tested on two impared and two unimpaired subjects. In order to check the exercise of the subjects’ upper limb muscles, muscle activations, heart rate and the arm range of movement during the different exercises have been
experimentally measured. The results showed that while exercising with the MOVit Lite system the muscle activation of the upper limb targeted muscles exceeded 20% of the Maximum Voluntary Contraction for one Becker Muscular Dystrophy subject. The other impaired participant reached a muscular activation of 87% of Maximum Voluntary Contraction the Deltoideus Posterior with a standard deviation of 32%. Regarding the metabolic cost, the interface does not require a high intensity of work as the mean heart rate during the activity was similar to the one in rest for all the participants. The range of motion of the arms obtained with the MOVit control mode was similar for unimpaired and impaired subjects, reaching up to 30 ± 7 cm for one unimpaired subject. The wrist rotation of the joystick control mode is also similar and reaches for one impaired subject up to 85° with a standard deviation of 6°. The feedback from the participants was positive marking the system as entertaining for daily use. The MOVit Lite system has fulfilled the main objectives stated at the beginning
of the project. Further tests with a larger sample of impaired subjects need to be carried out. Future steps also include the integration of this system in the powered MOVit wheelchair, which is an ongoing process. We are developing the hardware and control codes to get the same good performance as with the MOVit Lite system, developed in this master’s thesis
http://hdl.handle.net/2117/351315
ETSEIB-240.160300
Àrees temàtiques de la UPC::Ciències de la salut
Àrees temàtiques de la UPC::Enginyeria electrònica
Upper limb teleoperated control interface for powered wheelchair users
oai:upcommons.upc.edu:2117/808322023-05-19T09:52:48Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Fandos Jahrsetz, Alexander
author
2014
Interaction between humans and AGVs (Automatic guided vehicles) is often reduced to non-existent or just some pushbuttons, touchscreens or signalling lamps [17][18][19][20]. In recent years some researchers have attempted to improve the interaction through AGVs that follow people [53], path planning through gestures or robots that are able to ask for help if they cannot carry out a task on its own. [54] This project proposes extending the HRI (Human Robot Interaction) of Kivnon Logística's AGVs with gesture recognition and human localization through depth cameras and with emotion-simulating feedback through a screen. The aim for this technology introduction is to develop a shopping cart like AGV that follows people on gesture commands. The developed system will be built on Kivnon Logística's AGV model K11 mouse as a prototype. Later on a specific AGV will be designed. The final aim of this project is to introduce smart interactive AGVs in industrial environments focused in production areas.
http://hdl.handle.net/2117/80832
Àrees temàtiques de la UPC::Informàtica::Robòtica
Gesture based HRI for industrial AGVs
oai:upcommons.upc.edu:2117/987352020-02-12T20:43:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ramasubramanian, Darshan
author
2016-09-05
This work has two main objectives: the first one is identify the model of DC motors - using various techniques from experimental data; Secondly, we want to design appropriate control structures for this system, so as to meet the specifications requested. The second chapter describes the equipment with which we work throughout the project, consisting of two rotary brushed DC motors provided by the company Ingenia Motion Control. The third chapter describes various techniques of system identification that are used to characterize the motor dynamics, and applied to the motors provided by the company Ingenia Motion Control. And the final chapter covers the design of controllers plants based on the specification in bandwidth, and provides detailed structures of PID control and I-PD control.
This master thesis has been developed in the Institut de Robòtica i Informàtica Industrial (IRI) and in the company Ingenia Motion Control in the Firmware Division with the supervision of Vicenç Puig as director and Roger Juanpere and Rob Verhaart as advisers from the company.
http://hdl.handle.net/2117/98735
ETSEIB-240.119179
Àrees temàtiques de la UPC::Informàtica::Robòtica
Identification and control of DC motors
oai:upcommons.upc.edu:2117/3742772024-03-24T07:37:35Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Requena Gallego, José
author
2022-09-12
As more and more autonomous vehicles enter our road, new mechanisms must be considered to ensure the safe coordination between the autonomous vehicles. Al- though many algorithms have been proposed to coordinate autonomous vehicles, few of them have considered the robustness of the solution against disturbances. Therefore, in this master’s thesis, a vehicle coordination algorithm that uses vehicle to vehicle (V2V) communication is design in order to achieve collision free trajectories, while rejecting disturbances. Specifically, a robust tube-based model predictive control (MPC) scheme is proposed in order to control the autonomous vehicle. This controller uses series of zonotopic reachable sets (also known as tube) to compute a set of state and input constraints, which ensure the robust feasibility of the problem. To reduce the computational burden of the MPC opti- mization problem, the vehicle model is reformulated into a pseudo-linear model by transforming its non-linear equations into the linear parameter varying (LPV) form. The disturbance rejection is performed by a H∞-optimal corrective con- troller. Finally, the collision avoidance is achieved by a V2V coordination algo- rithm, in which the lateral bounds of a collision free path are computed. To validate the proposed control scheme, a series of simulations have been performed to test the disturbance rejection of the corrective controller, as well as the vehicle coordination capabilities. The results from these tests show that the proposed controller is effective in coordinating a multiple overtaking maneuver, while rejecting the disturbances.
http://hdl.handle.net/2117/374277
ETSEIB-240.171270
Àrees temàtiques de la UPC::Enginyeria mecànica::Disseny i construcció de vehicles::Automòbils
Safe Coordination of Autonomous Vehicles
oai:upcommons.upc.edu:2117/1676772019-09-23T10:05:05Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Giner Casas, Anna
author
2018
http://hdl.handle.net/2117/167677
Àrees temàtiques de la UPC::Enginyeria biomèdica::Robòtica mèdica
Ontology-based system for surgical workflow recognition
oai:upcommons.upc.edu:2117/3541812023-02-10T06:13:03Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Prieto Molina, Sergio
author
2021-10-21
Nowadays, drones use sensors such as GPS, gyroscopes, and accelerometers to obtain the position and use it to coordinate different autonomous vehicles. However, many times these sensors fail or obtain information affected by noise. Therefore the position of aerial vehicles is sometimes inaccurate, especially if drones fly where the GPS signal does not reach the drone’s GPS. Therefore, in situations where certain vehicles are coordinated, it is necessary to estimate their position by other methods. This thesis addresses the location problem by obtaining the position of the surrounding vehicles employing a stereoscopic camera. Furthermore, the location is constantly obtained by performing a continuous tracking of the vehicle. Thus, in this way, it is possible to provide a constant location of an object which has no GPS or inaccurate locating systems
http://hdl.handle.net/2117/354181
ETSEIB-240.160084
Àrees temàtiques de la UPC::Enginyeria electrònica
Àrees temàtiques de la UPC::Informàtica
Study for the localization of vehicles using a drone with stereoscopic vision
oai:upcommons.upc.edu:2117/1204632023-11-12T07:40:29Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Romero Ben, Luis
author
2018-06-21
The main focus of the project consists in the design of the control of a fuel-cell energy management system, so that it can be validated in simulation to finally implement it employing a FPGA-based structure. The project will be centered in the control system design, in which different techniques can be considered (LMIs, sliding...) and the FPGA utilization, as the controller will be programmed directly on the programmable device using HDL programming tools.
http://hdl.handle.net/2117/120463
ETSEIB-240.133766
FPGA-based control of fuel-cell based energy management systems
oai:upcommons.upc.edu:2117/984812020-02-12T20:43:40Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Verdezoto Pereira, Victor
author
2016-09-05
This master thesis aims to demonstrate the effectiveness of a fault detection strategy in the structure of the wind turbine through a new strategy involving singular spectrum analysis (SSA), statistical methods and methods based on frequency. Based on a dynamic model of wind turbine, the time series behavior was obtained in the first instance without presenting any system failure and a second instance of system failure. From these series we can intervene with SSA and with statistical methods (variance, means, covariance, Fisher criteria) to design the fault detection system. The baseline will be designed with 850 healthy samples and a total sampling time of 6.25 seconds. This baseline which will provide us with the components to be compared so that the system can detect various faults that occur (fault types: fixed value, gain factor, offset and dynamics changed) in an efficient way. The results show that a sensor fault system with a high percentage of effectiveness was designed. The greater or lesser effectiveness thereof will depend on the established base line and the components used.
http://hdl.handle.net/2117/98481
ETSEIB-240.119957
Àrees temàtiques de la UPC::Energies::Energia eòlica::Aerogeneradors
Àrees temàtiques de la UPC::Matemàtiques i estadística
Fault detection in structures (wind turbine) through statistical techniques, singular spectrum analysis (SSA) and frequency-based methods
oai:upcommons.upc.edu:2117/1710872019-12-11T08:18:26Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Castro Rios, Ramiro Saito
author
2018
This project is based on the researching of the deep neural network for the classification of amber gemstone seen as a great opportunity to expand the range of
application of the well known deep learning, which have been used more and more
in different fields. First, it is explained the theoretical concepts about machine learning and deep learning, besides a little comparison between them in order to for the
reader to make the thesis more understandable. It is also introduced the different
machine learning classifiers and the different methods of deep learning which have
been used during the presented work. After having the basic concepts it will be
explained the different methods that have been studied throughout the thesis to
find the best possible accuracy in the training, validation and testing of the data
set. Finally, it will be showcased the different results obtained followed by a short
explanation per each of them.
In this work the main technique that will be used for the researching is the method
of transfer learning. Due to the lack of huge amounts of database regarding the
amber stones, this technique can be a lot profitable to achieve the best accuracy.
Nevertheless other methods like training from scratch will be tested; in which it will
be used YOLO (You Only Looks Once). There is plenty of developed architectures
that can be used for transfer learning, but only AlexNet will be used, since is one of
the most used. The main tool for performing all the training and testing is MATLAB, which is quite suitable for processing images and with the latest version now
it is able to run deep neural network with ease.
http://hdl.handle.net/2117/171087
Àrees temàtiques de la UPC::Informàtica::Aplicacions de la informàtica
Researching of the deep neural network for amber gemstone classification
oai:upcommons.upc.edu:2117/3321272023-03-19T07:11:31Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Ramon Borràs, Marina
author
2020-11-12
For many people with upper limb disabilities, simple activities of daily living such as drinking, opening a door, or pushing an elevator button require the assistance of a caregiver; which reduces the independence of the individual. Assistive robotic systems controlled via human-robot interface could enable these people to perform this kind of tasks autonomously again and thereby increase their independence and quality of life. Moreover, this interface could encourage rehabilitation of motor functions because the individual would require to perform its remaining body movements and muscle activity to provide control signals. This project aims at developing a novel hybrid control interface that combines remaining movements and muscle activity of the upper body to control position and impedance of a robotic manipulator. This thesis presents a Cartesian position control system for KINOVA Gen3 robotic arm, which performs a proportional-derivative control low based to the Jacobian transpose method, that does not require inverse kinematics. A second control is proposed to change the robot’s rigidity in real-time based on measurements of muscle activity (sEMG). This control allows the user to modulate the robot’s impedance while performing a task. Moreover, it presents a body-machine interface that maps the motions of the upper body (head and shoulders) to the space of robot control signals. Its uses the principal component analysis algorithm for dimensionality reduction. The results demonstrate that combining the three methods presented above, the user can control robot positions with head and shoulders movements, while also adapting the robot’s impedance depending on its muscle activation. In the future work the performance of this system is going to be tested in patients with severe movement impairments.
http://hdl.handle.net/2117/332127
ETSEIB-240.153120
Àrees temàtiques de la UPC::Enginyeria mecànica
iMOVE: Development of a hybrid control interface based on sEMG and movement signals for an assistive robotic manipulator
oai:upcommons.upc.edu:2117/1673452020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Hu, Xiaoxuan
author
2018-04-27
The Simultaneous Localization and Mapping (SLAM) problem for mobile robots aims at build-ing a map of an unknown environment while simultaneously determining the robot’s position within this map. In the robotics community, SLAM is considered a solved problem in the most common settings, but an approach to the SLAM problem using minimal sensing information is still relevant both from the theoretical and practical point of view.
In this context, this M.Sc. thesis aims at studying and implementing a special type of SLAM solely based on the information provided by the on-board Inertial Measurement Unit (IMU) from a spherical mobile robot called Sphero, developed by Orbotix.
This IMU permits detecting impacts with the environment and performing odometry subject to significant errors. Thus, the SLAM problem will be restricted to environments defined by closed regions bounded by orthogonal walls meeting at convex and concave corners. In this kind of environments, the robot can follow a motion in contact with the walls, while detecting corners, till it considers that it has returned to the initial point. As a result, a map of the environment is completed.
We will show how it is possible to define the location of the corners or landmarks of the rectilinear environments from changes in the robot’s speed. A map representation, that consist of the location of the landmarks, their uncertainty, and their connection through straight walls, is defined. Three uncertainty models have been used and compared: a probabilistic model and two bounded-error models. Two basic operations have been implemented for the three uncertainty models: propagation and fusion.
Propagation is the basic operation that permits assigning an uncertainty to the location of a landmark, given the location and uncertainty of the previously visited landmark and the estimated error committed in the odometry. If this new landmark has already been visited, it has a previous assigned uncertainty. This uncertainty has then to be updated given the new uncertainty using a fusion operation. This concept of fusion and propagation of uncertainty applies to the three mentioned methods, they only differ in their implementation.
The probabilistic approach has dominated much of the work on low-level sensing processes handling uncertainty. Although probabilistic models which assume a uniform distribution inside a range have been used, the computational tractability of low-level sensing processes, under this approach, requires the general assumption that the experimental error is simply an additive term with Gaussian distribution, and the fusion operation is essentially that of maximum likelihood estimation. Nevertheless, it is difficult to give complete error analysis because the complexity of the process of extracting low level data. Instead, when using bounded-error models, every measurement is assumed to lead to an uncertainty set in the space of parameters where the actual value is bound to be, and the fusion operation essentially reduces to find a bound for the intersection of two sets.
We will show how the two bounded-error models are of interest in those situations in which no probabilistic description of errors is available, only bound on them are known. Thus, these models are of particular interest in minimalist robotics. Actually, one important goal of this thesis has been to show how, using simple sensor information, it is possible to solve complex tasks, such as computing a map of the boundaries of an office-like environment. Thus, we have tried to answer the basic question raised by minimalist robotics: How simple can a robot be while nevertheless accomplishing interesting tasks? There are at least three important reasons to focus on minimalist robots: (1) it encourages us to find what is the least amount of information needed to solve a certain task, giving insights into the task’s inherent complexity;(2) it also encourages us to use inexpensive sensors providing very limited or noisy sensing; and, finally, (3) it may allow us to manufacture inexpensive robots with low energy consumption. This thesis is structured as follows. Chapters 2 and 3 describe the basic theory behind the mapping solutions relying, respectively, on the Kalman filter an on the set theoretic approaches. Then, Chapter 4 describes the hardware and software of the Sphero robot, including the robot’s control, motion and calibration. Chapter 5 presented the implementation and the experimental results. Finally, Chapter 6 includes the conclusions and indicate possible extensions of this work.
http://hdl.handle.net/2117/167345
ETSEIB-240.131253
Àrees temàtiques de la UPC::Informàtica::Robòtica
SLAM with SPHEROS
oai:upcommons.upc.edu:2099.1/237022021-04-14T07:12:13Zcom_2099.1_4237com_2117_28592col_2099.1_13038
00925njm 22002777a 4500
dc
Luna Pacho, Julio Alberto
author
2014-06
A series of nonlinear state observers are developed to estimate unmeasurable variables in a Proton
Exchange Membrane Fuel Cell (PEMFC) system. A study of the most relevant variables
to estimate which will improve the e ciency of the system and its performance is done. The
implementation of the observers is based on the partial di erential equations that de ne the
dynamics of the plant, represented by a distributed parameters model. Forward and backward
discretization of the distributed model are performed to take advantage of the boundary conditions
of the problem. A super-twisting corrective control action is implemented in the nonlinear
observers to reduce the estimation error to zero in a finite amount of time. Simulation results
are presented to show the performance of the observers to recover the values of the gas species
concentrations and to extract conclusions for future research work.
http://hdl.handle.net/2099.1/23702
Àrees temàtiques de la UPC::Informàtica::Automàtica i control
Àrees temàtiques de la UPC::Energies::Tecnologia energètica::Emmagatzematge i transport de l’energia
State Observers Design for PEMFC Systems
oai:upcommons.upc.edu:2117/1668942020-02-12T20:43:36Zcom_2099.1_4237com_2117_28592col_2099.1_13038openAccess
00925njm 22002777a 4500
dc
Garcia Peroy, Xavier
author
2019-07-04
This project aims to design and implement a solution for a complex mobile manipulation task
using PAL Robotics’ robot TIAGo. The goal is to use the platform as an assistive robot, making
it capable of picking a desired object from a certain location and moving it to another. More
precisely the task in which the project will focus will be to pick a certain soda can from a table
with several of them on it and pour its contents into a glass placed on another table for the user
to have his/her drink. For that, all the perception, interaction, planning and mobile capabilities
of the robotic platform will be exploited in order to develop a suitable and complete solution.
The solution generated is modular so it can be launched as a whole process to complete the
entire challenge but also as independent modules to solve different tasks. This makes it possible
to easily integrate them to other processes in order to complete other similar tasks, making the
packages more versatile and adaptable.
The work has been divided between two master thesis projects. The first one focuses on developing the packages in charge of navigation and arm manipulation, which is reflected on this
memory. The second one, carried out by Miller Stiven Espinosa Muñoz, covers the perception
and task management blocks.
For the navigation part a solution has been implemented to map the environment and define
all the necessary poses to complete the manipulation process. The method developed links the
information of the map with visual marks placed on the tables of the scenario. In this way
the poses are defined with respect to these marks and linked to the map automatically. For
navigation itself the ROS 2D Navigation stack has been exploited to control the base and move it
successfully.
For the arm manipulation two different techniques, aside of the ones already built in the robot,
have been implemented. A technique based on generating trajectories using the geometric planner RRTConnect has been developed for the movements involved in picking the cans from the
table using the interface of The Kautham Project, a software for robot motion planning. Another
imitation method has been developed using the learning framework of Dynamic Movement Primitives to perform the movements involved in serving the soda drink. Both methods have been
implemented into a single package with functions capable of performing the trajectories by interfacing the services already built in the robot.
All the solutions developed have been proven to work properly to achieve the goal established.
The robot has been capable of mapping and navigating the room successfully while correctly
defining all the required poses around the tables to complete the task demanded. The arm
manipulation capabilities have been exploited to perform all the necessary movements to pick
and serve the soda cans.
The solution presents some flaws in terms of reliability and robustness mainly due to errors in
sensing the position of the cans and in localizing the robot in the map. Future work is suggested
in order to solve these problems and improve the performace of the system.
http://hdl.handle.net/2117/166894
ETSEIB-240.144756
Àrees temàtiques de la UPC::Informàtica
Mobile manipulation with the TiaGo robot: navigation and planning.
marc///col_2099.1_13038/100