http://hdl.handle.net/2117/3501 2013-05-21T23:06:27Z http://hdl.handle.net/2117/19089 Title: MUSS, a modular simulation system Authors: Guasch Petit, Antonio Abstract: In contrast with classical simulation languages, the present trends are evolving towards fully integrated interactive modelling and simulation evironmens. These environments have to combine interdisciplianry tecniques such as expert systems, object oriented programming and data base management.To achieve the above objectives, the architecture of the simulation programming language and that of the run-time simulation environment which exercises the models should be deinged allowing modularity and flexibility. Furthermore, the robustness of the environment should be reinforced. In this paper the MUSS simulation system is presented, emphasizing the innovative conepts: the hierachcical architecture of the MUSS simulation language, the preprocessor analysis anf segmentation pases and the structure of the run-time simulation environment. 2013-05-03T10:02:23Z http://hdl.handle.net/2117/19088 Title: Object-oriented continuous system simulation Authors: Guasch Petit, Antonio; Huntsinger, R C Abstract: Continuous model simulation systems have evolved in the present decade from the monolitic CSSL 1967 architecture to new and more powerful modular ones. However, in spite of this constant evolution, the majority of the continuous model simulation systems available in the market have an architecture based on the 1967 standard. Furthermore, the programming language in which the majority of theses systems are developed is still Fortran. In spite of the engineering value, the use of Fortran or other classical procedural languages has constrained the degreee of moduarity, felicibility and reliability of the simulation systems. Therefore, the adoptin of new protramming methodologies is advisable. At present, it seems that de object-oriented approach is one of the most promising. 2013-05-03T09:19:20Z http://hdl.handle.net/2117/19087 Title: Continuous simulation in Smalltalk Authors: Guasch Petit, Antonio Abstract: Although object-oriented languages such as Smalltalk do not seem very suitable for continuous simulation, they offer a set of characteristics and programming environments that are excellent for prototyping.This paper describes the previous author's experience developing continuous model simulation environments using FORTRAN and C programming languages and the current results obtained using Smalltalk. The transition from FORTRAN to C, from C to C++ and later on, the move to Smalltalk has been propelled by the need to develop more flexible and easy to use simulation environments.A continuous modeling and simulation environment has been implemented in the SmallTalk\V 286 system. The simulation environment , called SIMBIOS, strongly supports hierarchical modeling and simulation. In this system, users can examine their models in a highly flexible environment. Since interactive simulation is supported, the usual sequence of editing, translting, compiling, linking and simulating the model is avoided. 2013-05-03T09:08:44Z http://hdl.handle.net/2117/18756 Title: Precompiled submodels: A general sorting procedure Authors: Guasch Petit, Antonio; Hubert, R M Abstract: Simulation languages supporting hierarchical submodel structures, face sorting problems not present in monolithic CSSL67 models. These problems , result from algebraic relationships between submodels leading to information loops in the model code. This paper proposes a global solution, based on the segmentation of the submodels through graph analysis, and its promising application to sorting problems. 2013-04-10T09:10:46Z http://hdl.handle.net/2117/18754 Title: Towards a specification of the structure for continuous system simulation languages Authors: Guasch Petit, Antonio; Huber, RM Abstract: Present CSSIs are a valuable tool in systems engineering but some changes in the structure and a standardization of the language constructs would be useful. A comparative review of the structure in recent simulation ot piecewise continuous systems whose aim is to provide a modular way to build-up the model as well as to ease the experimentation, validation and debugging. 2013-04-10T08:56:14Z http://hdl.handle.net/2117/18753 Title: SIMBIOS: SIMulation Based on Icons and ObjectS Authors: Guasch Petit, Antonio; Luker, P Abstract: An interactive simulation environment prototype for the study of continuous models has been implemented using the SmallTalk/V system. Models are designed in a graphical manner and can be directly executed without processing, compiling and linking. Therefore, models can instantaneously react to new configurations. Owing to the setmentation of the submodel code, the system suports hierarchical simulation design and execution without restrictions. A directed graph, which reflects the computational flow, is used to direct the interactive experiments. 2013-04-10T08:25:23Z http://hdl.handle.net/2117/18482 Title: Submodels, models, experiments and projects Authors: Guasch Petit, Antonio; Monfort, X Abstract: SIMBIOS (SIMulation Based on Icons and Objects) is an object oriented continuous simulation environment written in Smalltalk/V Windows Models are defined graphically on the screen using a block diagram format. In this paper, the project concept is introduced. It is the element responsible of the simulation project management. A project is composed of a set of submodels, models and experiment. The submodels are grouped into models and every model has one or more experiments associated to it. A submodel can be simultaneously present in more than one model and any change made to the parameters of that submodel will be automatically seen in all the models where the submodel is present. This capability increases the flexibility of the simulation environment. 2013-03-22T11:05:57Z http://hdl.handle.net/2117/18469 Title: Integrating bus stop modeling and simulation in public transport network design: a survey Authors: Codina Sancho, Esteve; Fonseca Casas, Pau; Colls, Màxim; Montero Mercadé, Lídia Abstract: In this paper a survey of public transportation network design and assignment models is presented with the purpose to analyze the level of detail that is considered regarding aspects such as congestion and interaction between bus lines and passengers at stops. After briefly reviewing the route design models in which these aspects are taken into account, a review of transit assignment models and frequency setting models is made in which these factors are considered. Simulation approaches oriented to the evaluation of complex configurations of bus stops used as evaluation tools for the design and planning of bus lines are also described. Because of its relevance, special attention is paid to the so called Advanced Public Transportation Systems. 2013-03-22T09:55:28Z http://hdl.handle.net/2117/18367 Title: Real time diagnosis for a large gas turbine based on a deep model of the controller Authors: Guasch Petit, Antonio; Sarrate Estruch, Ramon Abstract: The ESPRIT project 6862 TIGER addresses the development of a real-time knowledge based diagnosis system for two gas turbines: a large scale industrial gas turbine and an auxiliary power unit for aviation. The project will combine the qualitative model based with the rule based diagnosis approach. This paper describes the rule based approach taken in the development of a real-time diagnostic system for a large scale gas turbine. Deep models will be automatically extracted from the deterministic knowledge embedded within the controller ladder logic diagrams. Heuristic knowledge will be extracted from the domain expert and engineering manuals. 2013-03-18T12:30:17Z http://hdl.handle.net/2117/18018 Title: Building optisim, optimizador NZEB Authors: Fonseca Casas, Pau; Colls, Màxim; Casanovas Garcia, Josep; Fonseca Casas, Antoni; Garrido Soriano, Núria Abstract: La simulación de la demanda energética de un edificio es una tarea compleja debido principalmente a dos factores clave. En primer lugar, es necesario definir la física y las complejas relaciones entre todos los elementos que pertenecen al modelo. En segundo lugar, es necesario entender estas relaciones por un equipo que normalmente está compuesto por personal con experiencia y formación diferente. En este trabajo, proponemos el uso de un lenguaje formal para representar el comportamiento dinámico de un modelo que representa todo el ciclo de vida de un edificio (diseño, construcción, uso‐mantenimiento y deconstrucción), y técnicas de co‐simulación para combinar motores de simulación diferentes y obtener lo mejor de cada uno. Se presenta un modelo de sostenibilidad, Building OptiSim, de tal manera que permite obtener una solución óptima. El edificio que utilizamos para realizar la simulación es la construcción LOW 3, proyecto presentado en el concurso Solar Decathlon 2010, por la ETSAV (UPC). 2013-02-28T11:46:56Z http://hdl.handle.net/2117/17567 Title: Using simulation to provide alternative solutions to the flowshop sequencing problem Authors: Juan Pérez, Angel Alejandro; Guix, Antoni; Adelantado Freixer, Ferran; Fonseca Casas, Pau; Ruiz García, Rubén Abstract: In this paper we present SS-GNEH, a simulation-based algorithm for the Permutation Flowshop Sequencing Problem (PFSP). Given a PFSP instance, the SSGNEH algorithm incorporates a randomness criterion to the classical NEH heuristic and starts an iterative process in order to obtain a set of alternative solutions, each of which outperforms the NEH algorithm. Thus, a random but oriented local search of the space of solutions is performed, and a list of "good alternative solutions" is obtained. We can then consider several desired properties per solution other than maximum time employed, such as balanced idle times among machines, number of completed jobs at a given target time, etc. This allows the decision-maker to consider multiple solution characteristics other than just those defined by the aprioristic objective function. Therefore, our methodology provides flexibility during the sequence selection process, which may help to improve the scheduling process. Several tests have been performed to discuss the effectiveness of this approach. The results obtained so far are promising enough to encourage further developments on the algorithm and its applications in real-life scenarios 2013-02-04T12:33:12Z http://hdl.handle.net/2117/17234 Title: SIM-Randsharp: A hybrid algorithm for solving the ARC routing problem with stochastic demands Authors: González, Sergio; Riera, Daniel; Juan, Angel A.; Elizondo, Mónica G.; Fonseca Casas, Pau Abstract: This paper proposes a new hybrid algorithm for solving the Arc Routing Problem with Stochastic Demands (ARPSD). Our approach combines Monte Carlo simulation (MCS) with the RandSHARP algorithm, which is designed for solving the Capacitated Arc Routing Problem (CARP) with deterministic demands. The RandSHARP algorithm makes use of a CARP-adapted version of the Clarke and Wright Savings heuristic, which was originally designed for the Vehicle Routing Problem. The RandSHARP algorithm also integrates a biased-randomized process, which allows it to obtain competitive results for the CARP in low computational times. The RandSHARP algorithm is then combined with MCS to solve the ARPSD. Some numerical experiments contribute to illustrate the potential benefits of our approach. 2013-01-09T12:32:23Z http://hdl.handle.net/2117/17226 Title: NZEB. Optimización mediante co-simulación. Estudio de caso (e)Co, prototipo del concurso Solar Decathlon 2012 Authors: Fonseca Casas, Pau Abstract: Actualmente, en los estudios de arquitectura, hay una necesidad de integrar diferentes disciplinas y técnicos especializados en el grupo de trabajo; donde se hace imprescindible el uso de simuladores y software CAD (Computer Aided Design) para optimizar el proyecto en fase de diseño a nivel energético y ambiental. Normalmente, los arquitectos e ingenieros gastan mucho tiempo y recursos en levantar el modelo en formato BIM (Building Information Modeling) para ser después analizado en motores de cálculo como EnergyPlus o Doe-2; simplemente llegando a realizar unos cuantos experimentos concretos del modelo de estudio; dejando sin analizar muchas otras posibilidades y sin llegar a optimizarlo ni obtener la mejor configuración del proyecto; de tal manera que, por falta de tiempo, solamente se analiza y se hace un estudio parcial, y muchas veces insuficiente, para saber si la configuración escogida ha sido la más eficiente a nivel energético y económico. La presente publicación muestra las posibilidades de la co-simulación, la cual da respuesta a las dos cuestiones planteadas anteriormente: Para implementar el modelo de co-simulación se utiliza el lenguaje Specification and Description Language (SDL), lenguaje formal gráfico que permite fácilmente integrar y combinarse con otros lenguajes, solucionando la problemática de la integración de diferentes áreas de especialidad. De tal manera que, el SDL, por un lado, es capaz de definir los diferentes modelos gracias a un sistema de diagramas, mucho más ágil y fácil de ser legible que otros lenguajes de programación y capaz de ser entendido fácilmente por todos los integrantes del equipo, arquitectos e ingenieros en un conjunto; y por el otro, nos permite tener definido un modelo capaz de integrar nuevos procesos de cálculo y combinarse con otros simuladores de forma directa. En respuesta al segundo punto, el co-simulador utiliza diferentes motores de cálculo para realizar la optimización del sistema, de forma autónoma y sin tener que entrar y detallar cada experimento concretamente. El co-simulador, mediante la base de datos integrada, optimiza el modelo en base a diferentes soluciones constructivas (pudiendo determinar nuevas configuraciones), orientaciones, zonas climáticas, sistemas de climatización activos y formas geométricas. Es capaz de combinar todas las situaciones deseadas para cada experimento encontrando la solución óptima del conjunto. El co-simulador vía cloud, integrado en página web para el usuario, ha sido utilizado para optimizar, a nivel de demanda energética (optimización mediante diferentes soluciones constructivas), el proyecto (e)CO, prototipo del concurso Solar Decathlon del 2012, como ejemplo de edificación sostenible del grupo de la Escuela Técnica Superior de Arquitectura del Vallés (ETSAV – UPC). El objetivo del estudio es proponer un nuevo sistema de co-simulación que resuelve las problemáticas planteadas anteriormente, abriendo nuevos campos y mejorando en gran medida las posibilidades de estudio en la fase de proyecto para atender en la totalidad la complejidad de un sistema arquitectónico. El co-simulador es un proyecto de investigación de al UPC, que está en desarrollo continuo integrando nuevas funcionalidades y más potencia de cálculo, con la intención de abrirse a muchos otros campos tecnológicos. El informe final del caso de estudio, revela la gran utilidad del software y el enorme potencial que posee para ser aplicado en multitud de casos. Gracias a la potencia de cálculo y después de simular más de 4000 situaciones, se ha podido determinar, con exactitud, la configuración idónea para el caso concreto del proyecto (e)CO. 2013-01-09T10:28:36Z http://hdl.handle.net/2117/15525 Title: Modelling surgical pavilions and a unit of anaesthesia on a Chilean hospital using Specification and Description Language Authors: Leiva Olmos, J.; Fonseca Casas, Pau; Ocaña, J. Abstract: This work addresses the problem of performing a formal modelling of the processes related to the surgical pavilions and an anaesthesia unit on a Chilean hospital. To perform this modelling we used Specification and Description Language (SDL). The model was very successful in order to document and to understand the tacit knowledge of the unit and facilitates the simulation. 2012-03-08T11:39:14Z http://hdl.handle.net/2117/14519 Title: Threshold volatility models: forecasting performance Authors: Marquez, M. D.; Muñoz Gracia, María del Pilar; Marti, Manuel; Acosta Argueta, Lesly María Abstract: The aim of this paper is to compare the forecasting performance of competing volatility models, in order to capture the asymmetric effect in the volatility. We focus on examining the relative out-of-sample forecasting ability of the models (SETAR-TGARCH and SETAR-THSV), which contain the introduction of regimes based on thresholds in the mean equation and volatility equation, compared to the GARCH model and SV model. For each model, we consider two cases: Gaussian and t-Student measurement noise distribution. An important problem when evaluating the predictive ability of volatility models is that the “true” underlying process is not observable and thus a proxy must be defined for the unobservable volatility. To attain our proposal, the proxy volatility measure and the loss function must also be decided to ensure a correct ranking of models. Our empirical application suggests the following results: when time series include leverage effects on the mean, the introduction of threshold in the mean and variance equations produces more accurate predictions. If the leverage in the mean is not important, then the SVt is flexible enough to beat the threshold models. 2012-01-13T11:49:49Z