Departament d'Enginyeria de Sistemes, Automàtica i Informàtica Industrial
http://hdl.handle.net/2117/3929
2024-03-28T14:53:30Z
2024-03-28T14:53:30Z
Anomaly detection in gas turbines using outlet energy analysis with cluster-based matrix profile
Bagherzade Ghazvini, Mina
Sànchez-Marrè, Miquel
Naderi, Davood
Angulo Bahón, Cecilio
http://hdl.handle.net/2117/405434
2024-03-27T09:51:19Z
2024-03-27T09:48:17Z
Anomaly detection in gas turbines using outlet energy analysis with cluster-based matrix profile
Bagherzade Ghazvini, Mina; Sànchez-Marrè, Miquel; Naderi, Davood; Angulo Bahón, Cecilio
Gas turbines play a key role in generating power. It is really important that they work efficiently, safely, and reliably. However, their performance can be adversely affected by factors such as component wear, vibrations, and temperature fluctuations, often leading to abnormal patterns indicative of potential failures. As a result, anomaly detection has become an area of active research. Matrix Profile (MP) methods have emerged as a promising solution for identifying significant deviations in time series data from normal operational patterns. While most existing MP methods focus on vibration analysis of gas turbines, this paper introduces a novel approach using the outlet power signal. This modified approach, termed Cluster-based Matrix Profile (CMP) analysis, facilitates the identification of abnormal patterns and subsequent anomaly detection within the gas turbine engine system. Significantly, CMP analysis not only accelerates processing speed, but also provides user-friendly support information for operators. The experimental results on real-world gas turbines demonstrate the effectiveness of our approach in the early detection of anomalies and potential system failures.
2024-03-27T09:48:17Z
Bagherzade Ghazvini, Mina
Sànchez-Marrè, Miquel
Naderi, Davood
Angulo Bahón, Cecilio
Gas turbines play a key role in generating power. It is really important that they work efficiently, safely, and reliably. However, their performance can be adversely affected by factors such as component wear, vibrations, and temperature fluctuations, often leading to abnormal patterns indicative of potential failures. As a result, anomaly detection has become an area of active research. Matrix Profile (MP) methods have emerged as a promising solution for identifying significant deviations in time series data from normal operational patterns. While most existing MP methods focus on vibration analysis of gas turbines, this paper introduces a novel approach using the outlet power signal. This modified approach, termed Cluster-based Matrix Profile (CMP) analysis, facilitates the identification of abnormal patterns and subsequent anomaly detection within the gas turbine engine system. Significantly, CMP analysis not only accelerates processing speed, but also provides user-friendly support information for operators. The experimental results on real-world gas turbines demonstrate the effectiveness of our approach in the early detection of anomalies and potential system failures.
Influence of the fuzzy function on the estimation of the fuzzy sample entropy with fixed tolerance values for the evaluation of surface EMG muscle activity
Torres Cebrián, Abel
Estrada Petrocelli, Luis Carlos
http://hdl.handle.net/2117/405096
2024-03-25T01:37:32Z
2024-03-21T13:57:37Z
Influence of the fuzzy function on the estimation of the fuzzy sample entropy with fixed tolerance values for the evaluation of surface EMG muscle activity
Torres Cebrián, Abel; Estrada Petrocelli, Luis Carlos
Fixed sample entropy (fSampEn) is a technique that has demonstrated superior performance to other amplitude estimators for assessing respiratory muscle electromyographic activity. This technique is based on the calculation of sample entropy (SampEn) using fixed tolerance thresholds. Fuzzy entropy (FuzzyEn) introduces an improvement to the SampEn algorithm based on the use of a fuzzy measure to evaluate the similarity between vectors. However, several fuzzy functions have been used to calculate the FuzzyEn, and not all of them allow an effective comparison with the SampEn calculation parameters. In the present work, an analysis of the different fuzzy functions previously used has been carried out and a new sigmoid fuzzy function for the calculation of FuzzyEn with fixed tolerance thresholds (fFuzzyEn) has been proposed. The results show that the proposed fuzzy function outperformed both fSampEn and previously proposed FuzzyEn-based algorithms. These results suggest that fFuzzyEn could improve the assessment of muscle activity providing potentially useful diagnostic information.
2024-03-21T13:57:37Z
Torres Cebrián, Abel
Estrada Petrocelli, Luis Carlos
Fixed sample entropy (fSampEn) is a technique that has demonstrated superior performance to other amplitude estimators for assessing respiratory muscle electromyographic activity. This technique is based on the calculation of sample entropy (SampEn) using fixed tolerance thresholds. Fuzzy entropy (FuzzyEn) introduces an improvement to the SampEn algorithm based on the use of a fuzzy measure to evaluate the similarity between vectors. However, several fuzzy functions have been used to calculate the FuzzyEn, and not all of them allow an effective comparison with the SampEn calculation parameters. In the present work, an analysis of the different fuzzy functions previously used has been carried out and a new sigmoid fuzzy function for the calculation of FuzzyEn with fixed tolerance thresholds (fFuzzyEn) has been proposed. The results show that the proposed fuzzy function outperformed both fSampEn and previously proposed FuzzyEn-based algorithms. These results suggest that fFuzzyEn could improve the assessment of muscle activity providing potentially useful diagnostic information.
Non-intrusive condition monitoring based on event detection and functional data clustering
Bermeo Ayerbe, Miguel Ángel
Ocampo-Martínez, Carlos
Díaz Rozo, Javier
http://hdl.handle.net/2117/405014
2024-03-25T01:34:52Z
2024-03-20T12:33:55Z
Non-intrusive condition monitoring based on event detection and functional data clustering
Bermeo Ayerbe, Miguel Ángel; Ocampo-Martínez, Carlos; Díaz Rozo, Javier
Implementing monitoring electricity consumption strategies in industrial environments provides improvements in both the maintenance process and energy efficiency. The contribution of this work is an industry-oriented non-intrusive load monitoring approach based on an unsupervised algorithm, encompassing a method of event detection, functional data clustering, and condition monitoring. With this method, multiple devices can be monitored by only one electric meter in industrial environments, enhancing the early detection of anomalies and energy inefficiencies. The proposed approach presents a robust event detection to deal with different industrial contexts due to its intuitive parameters, which allows adapting the detection to be more sensitive or to filter out higher noise variations. Unlike other feature-based clusterings, the proposed functional data clustering enables high-precision identification of transient state patterns, characterizing specific shapes for each pattern to properly cluster them, and provides higher reliability during load detection. Thus, this load detection segments the power consumption to extract transient states and identify which load acts on each event based on functional data clusters. By detecting when loads start to consume, the proposed energy disaggregation extracts the frequency spectrum of each device from the aggregate current consumption, which is used in a condition monitoring strategy to track the spectrum behavior of each device. In this way, the condition of multiple loads can be monitored using a single electric meter, whose information can be relevant to accurately schedule maintenance interventions and detect anomalies or inefficiencies early. The proposed approach was validated in three industrial contexts: The load detection accuracy was verified in an industrial testbed, giving a precision higher than 99% for monitoring five devices. The second and third industrial scenarios validate the accuracy of the proposed condition monitoring method. The last scenario was carried out at Bilbao airport to track the condition of multiple conveyor belts of a baggage handling system located at check-in. As a result, the degradation trend of three sets of conveyor belts was monitored.
2024-03-20T12:33:55Z
Bermeo Ayerbe, Miguel Ángel
Ocampo-Martínez, Carlos
Díaz Rozo, Javier
Implementing monitoring electricity consumption strategies in industrial environments provides improvements in both the maintenance process and energy efficiency. The contribution of this work is an industry-oriented non-intrusive load monitoring approach based on an unsupervised algorithm, encompassing a method of event detection, functional data clustering, and condition monitoring. With this method, multiple devices can be monitored by only one electric meter in industrial environments, enhancing the early detection of anomalies and energy inefficiencies. The proposed approach presents a robust event detection to deal with different industrial contexts due to its intuitive parameters, which allows adapting the detection to be more sensitive or to filter out higher noise variations. Unlike other feature-based clusterings, the proposed functional data clustering enables high-precision identification of transient state patterns, characterizing specific shapes for each pattern to properly cluster them, and provides higher reliability during load detection. Thus, this load detection segments the power consumption to extract transient states and identify which load acts on each event based on functional data clusters. By detecting when loads start to consume, the proposed energy disaggregation extracts the frequency spectrum of each device from the aggregate current consumption, which is used in a condition monitoring strategy to track the spectrum behavior of each device. In this way, the condition of multiple loads can be monitored using a single electric meter, whose information can be relevant to accurately schedule maintenance interventions and detect anomalies or inefficiencies early. The proposed approach was validated in three industrial contexts: The load detection accuracy was verified in an industrial testbed, giving a precision higher than 99% for monitoring five devices. The second and third industrial scenarios validate the accuracy of the proposed condition monitoring method. The last scenario was carried out at Bilbao airport to track the condition of multiple conveyor belts of a baggage handling system located at check-in. As a result, the degradation trend of three sets of conveyor belts was monitored.
Neuro-fuzzy Takagi Sugeno observer for fault diagnosis in wind turbines
Pérez Pérez, Esvan de Jesús
Puig Cayuela, Vicenç
López Estrada, Francisco Ronay
Valencia Palomo, Guillermo
Santos Ruiz, Ildeberto
http://hdl.handle.net/2117/405005
2024-03-25T01:42:49Z
2024-03-20T10:32:42Z
Neuro-fuzzy Takagi Sugeno observer for fault diagnosis in wind turbines
Pérez Pérez, Esvan de Jesús; Puig Cayuela, Vicenç; López Estrada, Francisco Ronay; Valencia Palomo, Guillermo; Santos Ruiz, Ildeberto
This work proposes a method for fault diagnosis based on Takagi Sugeno (TS) observers and convex models identified with a multioutput adaptive neuro-fuzzy inference system (MANFIS) derived from structural analysis. A bank of zonotopic TS observers is implemented to detect sensors and actuators faults. Unlike other works that require data from fault scenarios to train the MANFIS neural network, only fault-free data are considered. In addition, uncertainty related to aerodynamic loads and measurement noise is considered for testing the proposed method's robustness. The method performance is evaluated using measurements from a 5 MW wind turbine benchmark.
2024-03-20T10:32:42Z
Pérez Pérez, Esvan de Jesús
Puig Cayuela, Vicenç
López Estrada, Francisco Ronay
Valencia Palomo, Guillermo
Santos Ruiz, Ildeberto
This work proposes a method for fault diagnosis based on Takagi Sugeno (TS) observers and convex models identified with a multioutput adaptive neuro-fuzzy inference system (MANFIS) derived from structural analysis. A bank of zonotopic TS observers is implemented to detect sensors and actuators faults. Unlike other works that require data from fault scenarios to train the MANFIS neural network, only fault-free data are considered. In addition, uncertainty related to aerodynamic loads and measurement noise is considered for testing the proposed method's robustness. The method performance is evaluated using measurements from a 5 MW wind turbine benchmark.
Measuring high-resolution sleep position in adolescents over 4 nights with smartphone accelerometers
Castillo Escario, Yolanda
Blanco Almazán, Dolores
Ferrer Lluís, Ignasi
Jané Campos, Raimon
http://hdl.handle.net/2117/404986
2024-03-25T01:39:54Z
2024-03-20T07:55:45Z
Measuring high-resolution sleep position in adolescents over 4 nights with smartphone accelerometers
Castillo Escario, Yolanda; Blanco Almazán, Dolores; Ferrer Lluís, Ignasi; Jané Campos, Raimon
Sleep position affects sleep quality and the severity of different diseases. Classical methods to measure sleep position are complex, expensive, and difficult to use outside the laboratory. Wearables and smartphones can help to address these issues to track sleep position at home over several nights. In this study, we monitor high-resolution sleep position in 13 adolescents over 4 nights using smartphone accelerometer data. We aim to investigate the distribution of sleep positions and position changes in adolescents, study their variability across nights, and propose new measures related to nocturnal body movements. We developed a new index, the mean sleep angle change per hour, and calculated three other measures: position shifts per hour, mean time at each position, and periods of immobility. Our results indicate that participants spent 56% of the time on the side (32% right and 24% left), 32% in supine, and 12% in prone position, similar to what happens in adults. However, adolescents moved more than adults during sleep according to all measures. There was some variability between nights, but lower than the inter-subject variability. In conclusion, this work systematically analyzes sleep position over several nights in adolescents, a largely unstudied population, and offers innovative solutions and measures for high-resolution sleep position monitoring in a simple and cost-effective way.Clinical Relevance— Our study characterizes sleep position in adolescents and provides novel unobtrusive methods and quantitative indices to monitor high-resolution sleep position at home during multiple nights.
© 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting /republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works
2024-03-20T07:55:45Z
Castillo Escario, Yolanda
Blanco Almazán, Dolores
Ferrer Lluís, Ignasi
Jané Campos, Raimon
Sleep position affects sleep quality and the severity of different diseases. Classical methods to measure sleep position are complex, expensive, and difficult to use outside the laboratory. Wearables and smartphones can help to address these issues to track sleep position at home over several nights. In this study, we monitor high-resolution sleep position in 13 adolescents over 4 nights using smartphone accelerometer data. We aim to investigate the distribution of sleep positions and position changes in adolescents, study their variability across nights, and propose new measures related to nocturnal body movements. We developed a new index, the mean sleep angle change per hour, and calculated three other measures: position shifts per hour, mean time at each position, and periods of immobility. Our results indicate that participants spent 56% of the time on the side (32% right and 24% left), 32% in supine, and 12% in prone position, similar to what happens in adults. However, adolescents moved more than adults during sleep according to all measures. There was some variability between nights, but lower than the inter-subject variability. In conclusion, this work systematically analyzes sleep position over several nights in adolescents, a largely unstudied population, and offers innovative solutions and measures for high-resolution sleep position monitoring in a simple and cost-effective way.Clinical Relevance— Our study characterizes sleep position in adolescents and provides novel unobtrusive methods and quantitative indices to monitor high-resolution sleep position at home during multiple nights.
Integral sliding-mode fault-tolerant pitch control of wind turbines
Serrano, Fernando E.
Puig Cayuela, Vicenç
http://hdl.handle.net/2117/404934
2024-03-25T01:42:47Z
2024-03-19T12:44:02Z
Integral sliding-mode fault-tolerant pitch control of wind turbines
Serrano, Fernando E.; Puig Cayuela, Vicenç
In this paper, an integral sliding-mode fault-tolerant pitch control of wind turbines is presented. The proposed approach uses a fault diagnosis strategy which consists of a sliding-mode fault diagnosis observer. This observer is based on using an integral sliding-mode estimation scheme by using a suitable Lyapunov functional. Based on the previous fault diagnosis strategy, an integral sliding mode controller is designed by selecting an appropriate sliding mode surface in order to obtain the fault tolerant-control law obtained by also selecting appropriated Lyapunov functional. A wind-turbine case study is used to validate in simulation the the proposed approach.
2024-03-19T12:44:02Z
Serrano, Fernando E.
Puig Cayuela, Vicenç
In this paper, an integral sliding-mode fault-tolerant pitch control of wind turbines is presented. The proposed approach uses a fault diagnosis strategy which consists of a sliding-mode fault diagnosis observer. This observer is based on using an integral sliding-mode estimation scheme by using a suitable Lyapunov functional. Based on the previous fault diagnosis strategy, an integral sliding mode controller is designed by selecting an appropriate sliding mode surface in order to obtain the fault tolerant-control law obtained by also selecting appropriated Lyapunov functional. A wind-turbine case study is used to validate in simulation the the proposed approach.
Job shop scheduling with limited-capacity buffers using constraint programming and genetic algorithms
Pedrosa Alias, Javier
Puig Cayuela, Vicenç
http://hdl.handle.net/2117/404930
2024-03-25T01:42:02Z
2024-03-19T12:19:52Z
Job shop scheduling with limited-capacity buffers using constraint programming and genetic algorithms
Pedrosa Alias, Javier; Puig Cayuela, Vicenç
This article aims to propose a new approach for solving production planning and scheduling in the process industries, in such a way to be adaptable to any manufacturing plant and exploring the use of innovative AI-style technologies. The main contributions of the work are: (i) the design of a specific data format to describe any manufacturing plant (including resources, layout and production recipes), being the input of the method; and (ii) the consideration of limited-capacity production lines with intermediate and final buffers in the optimization. The method involves two stages: the first one corresponds to a deterministic optimization algorithm based on Constraint Programming modelling to solve the JSSP in an ideal scenario with no storage limitation; while the second one is a Genetic Algorithm that only comes into play when the solutions obtained from the first one are infeasible for the available storage, so it is a complementary layer to try to solve the mismatches stochastically.
2024-03-19T12:19:52Z
Pedrosa Alias, Javier
Puig Cayuela, Vicenç
This article aims to propose a new approach for solving production planning and scheduling in the process industries, in such a way to be adaptable to any manufacturing plant and exploring the use of innovative AI-style technologies. The main contributions of the work are: (i) the design of a specific data format to describe any manufacturing plant (including resources, layout and production recipes), being the input of the method; and (ii) the consideration of limited-capacity production lines with intermediate and final buffers in the optimization. The method involves two stages: the first one corresponds to a deterministic optimization algorithm based on Constraint Programming modelling to solve the JSSP in an ideal scenario with no storage limitation; while the second one is a Genetic Algorithm that only comes into play when the solutions obtained from the first one are infeasible for the available storage, so it is a complementary layer to try to solve the mismatches stochastically.
Robust tube-based TS-MPC for safe coordination of autonomous vehicle
Requena Gallego, José
Puig Cayuela, Vicenç
http://hdl.handle.net/2117/404923
2024-03-25T01:41:25Z
2024-03-19T11:59:30Z
Robust tube-based TS-MPC for safe coordination of autonomous vehicle
Requena Gallego, José; Puig Cayuela, Vicenç
In this work, a robust vehicle control scheme is proposed, which is capable of coordinating with nearby vehicles in order to optimally compute control actions that achieve collision-free overtaking maneuvers. The control actions are computed online by a global model predictive control (MPC) controller, which assumes a nominal disturbance-free vehicle model. To reduce the computational burden of the MPCs optimization problem, the vehicle model is reformulated into a pseudo-linear Takagi-Sugeno (TS) representation. Furthermore, the mismatch error between the real and the nominal model is corrected by a local TS H8-optimal state-feedback controller. Moreover, the robust feasibility of the MPCs optimization problem is guaranteed by implementing a tube-based architecture. Finally, the proposed control scheme is tested and validated in a high-fidelity simulation, in which the controlled vehicle was capable of overtaking multiple vehicles while rejecting disturbances.
2024-03-19T11:59:30Z
Requena Gallego, José
Puig Cayuela, Vicenç
In this work, a robust vehicle control scheme is proposed, which is capable of coordinating with nearby vehicles in order to optimally compute control actions that achieve collision-free overtaking maneuvers. The control actions are computed online by a global model predictive control (MPC) controller, which assumes a nominal disturbance-free vehicle model. To reduce the computational burden of the MPCs optimization problem, the vehicle model is reformulated into a pseudo-linear Takagi-Sugeno (TS) representation. Furthermore, the mismatch error between the real and the nominal model is corrected by a local TS H8-optimal state-feedback controller. Moreover, the robust feasibility of the MPCs optimization problem is guaranteed by implementing a tube-based architecture. Finally, the proposed control scheme is tested and validated in a high-fidelity simulation, in which the controlled vehicle was capable of overtaking multiple vehicles while rejecting disturbances.
Zonotopic set-membership state estimation for switched LPV systems
Zhang, Shuang
Puig Cayuela, Vicenç
Ifqir, Sara
http://hdl.handle.net/2117/404922
2024-03-25T01:40:46Z
2024-03-19T11:50:20Z
Zonotopic set-membership state estimation for switched LPV systems
Zhang, Shuang; Puig Cayuela, Vicenç; Ifqir, Sara
This paper addresses the state estimation problem for switched discrete-time Linear Parameter Varying (LPV) systems with mensurable and unmeasurable scheduling parameters. A zonotopic switched polytopic state estimator, considering parameter uncertainty, is proposed based on a Set-Membership Approach (SMA). Taking Average Dwell Time (ADT) into account, a new criterion is proposed to guarantee the convergence of the estimation. An application to vehicle lateral dynamics state estimation is used as case study. Simulation results reveal the effectiveness of the proposed algorithm and demonstrate advantages over the existing methods.
2024-03-19T11:50:20Z
Zhang, Shuang
Puig Cayuela, Vicenç
Ifqir, Sara
This paper addresses the state estimation problem for switched discrete-time Linear Parameter Varying (LPV) systems with mensurable and unmeasurable scheduling parameters. A zonotopic switched polytopic state estimator, considering parameter uncertainty, is proposed based on a Set-Membership Approach (SMA). Taking Average Dwell Time (ADT) into account, a new criterion is proposed to guarantee the convergence of the estimation. An application to vehicle lateral dynamics state estimation is used as case study. Simulation results reveal the effectiveness of the proposed algorithm and demonstrate advantages over the existing methods.
Nonlinear observer for online concentration estimation in vanadium flow batteries based on half-cell voltage measurements
Puleston, Thomas Paul
Cecilia Piñol, Andreu
Costa Castelló, Ramon
Serra, Maria
http://hdl.handle.net/2117/404918
2024-03-25T01:43:53Z
2024-03-19T11:14:50Z
Nonlinear observer for online concentration estimation in vanadium flow batteries based on half-cell voltage measurements
Puleston, Thomas Paul; Cecilia Piñol, Andreu; Costa Castelló, Ramon; Serra, Maria
This paper presents a nonlinear observer to estimate the active species concentrations in vanadium flow batteries. To conduct the estimation, the observer relies only on current, flow rate and two half-cell voltage measurements. In contrast to previous works in the field, the proposed observer is capable to deal simultaneously with two significant and challenging conditions: (1) a not necessarily high flow rate, which results in different concentrations for tanks and cells, and (2) presence of crossover and oxidation side reactions, that result in imbalance between the electrolytes on the positive and negative sides of the system. The stability and convergence of the observer are formally demonstrated using a Lyapunov analysis and subsequently validated through comprehensive computer simulations. Finally, utilising the information provided by the observer, a strategy to independently regulate the flow rate of each electrolyte based on their individual state of charge is developed.
2024-03-19T11:14:50Z
Puleston, Thomas Paul
Cecilia Piñol, Andreu
Costa Castelló, Ramon
Serra, Maria
This paper presents a nonlinear observer to estimate the active species concentrations in vanadium flow batteries. To conduct the estimation, the observer relies only on current, flow rate and two half-cell voltage measurements. In contrast to previous works in the field, the proposed observer is capable to deal simultaneously with two significant and challenging conditions: (1) a not necessarily high flow rate, which results in different concentrations for tanks and cells, and (2) presence of crossover and oxidation side reactions, that result in imbalance between the electrolytes on the positive and negative sides of the system. The stability and convergence of the observer are formally demonstrated using a Lyapunov analysis and subsequently validated through comprehensive computer simulations. Finally, utilising the information provided by the observer, a strategy to independently regulate the flow rate of each electrolyte based on their individual state of charge is developed.