Pasar al contenido principal

GESTION SOSTENIBLE Y BASADA EN APRENDIZAJE DE SISTEMAS MULTI-RECURSO DE GRAN ESCALA

PID2023-148840OB-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2023
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 40
Encontrada(s) 1 página(s)

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Linear quadratic control of nonlinear systems with Koopman operator learning and the Nyström method

  • Caldarelli, Edoardo
  • Chatalic, Antoine
  • Colomé, Adrià
  • Molinari, Cesare
  • Ocampo-Martínez, Carlos
  • Torras, Carme
  • Rosasco, Lorenzo
In this paper, we study how the Koopman operator framework can be combined with kernel methods to effectively control nonlinear dynamical systems. While kernel methods have typically large computational requirements, we show how random subspaces (Nystr\"om approximation) can be used to achieve huge computational savings while preserving accuracy. Our main technical contribution is deriving theoretical guarantees on the effect of the Nystr\"om approximation. More precisely, we study the linear quadratic regulator problem, showing that the approximated Riccati operator converges at the rate $m^{-1/2}$, and the regulator objective, for the associated solution of the optimal control problem, converges at the rate $m^{-1}$, where $m$ is the random subspace size. Theoretical findings are complemented by numerical experiments corroborating our results., E. Caldarelli, A. Colomé and C. Torras acknowledge support from the project CLOTHILDE (“CLOTH manIpulation Learning from DEmonstrations”), funded by the European Research Council(ERC) under the European Union’s Horizon 2020 research and innovation programme (Advanced Grant agreement No 741930). E. Caldarelli acknowledges travel support from ELISE (GA No 951847). C. Molinari is part of “GNAMPA” (INdAM) and has been supported by the projects MIUR Excellence Department awarded to DIMA UniGe CUP D33C23001110001, AFOSR FA8655-22-1-7034, MIUR-PRIN 202244A7YL and PON “Ricerca e Innovazione” 2014–2020. C. Ocampo-Martinez acknowledges the support from the project SEAMLESS (PID2023-148840OB-I00) funded by MCIN/AEI. L. Rosasco acknowledges the financial support of the European Research Council, (grant SLING 819789), the European Commission, (Horizon Europe grant ELIAS 101120237), the US Air Force Office of Scientific Research (FA8655-22-1-7034), the Ministry of Education, University and Research (FARE grant ML4IP R205T7J2KP; grant BAC FAIR PE00000013 funded by the EU - NGEU) and the Center for Brains, Minds and Machines (CBMM) ., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Neural network-based leak localization in water distribution networks using the gravity center of pressure measurements

  • Gómez-Coronel, Leonardo
  • Blesa, Joaquim
  • Santos-Ruiz, Ildeberto
  • López-Estrada, Francisco Ronay
  • Puig, Vicenç
A novel methodology for leak diagnosis in urban water distribution systems (WDS) is proposed. Small leaks are simulated using a well-calibrated EPANET model of the WDS. Considering only the known topology of the WDS, and pressure head values recorded at some nodes, the center of gravity of pressure is computed. Under nominal (leak-free) operation the position of the center of gravity varies predictably, but leaks cause variations on its position. Sensor-measurements with a duration of 24 h are used to compute residual coordinates from leak-free operation and used to train a LSTM neural network implemented in MATLAB for leak classification. Results are presented for the leak localization task considering two levels of resolution: identifying the general sector and pinpointing the specific node where the leak occurs. Tests are performed on a benchmark and real-world WDS obtaining a good performance with simulated data under steady-state and variable demand conditions. The impact of measurement noise is addressed by including the measured outflow from the reservoir as a third dimension to the training data., This work was developed within the framework of RICCA “Red Internacional de Control y Cómputo Aplicados”. Thanks to CONAHCYT and Tecnológico Nacional de México for the funding granted for this research through project 20,212.24-P. We would also like to thank the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE for supporting this research., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Techno-economic analysis of a renewable hydrogen refuelling station network: The case study of Spain

  • García Mínguez, Beatriz
  • Cardona, Pol
  • Valiño García, Luis
  • Ocampo-Martínez, Carlos
  • Siemon, Lukas
  • Serra-Prat, María
11 figures, 16 tables.-- Under a Creative Commons BY-NC-ND 4.0 license., Hydrogen-powered Fuel Cell Electric Vehicles (FCEVs) offer a promising pathway to reduce emissions of the transport sector, particularly when fuelled by renewable hydrogen produced through water electrolysis using Renewable Energy Sources (RESs). However, adoption is constrained by the limited refuelling infrastructure and the current higher readiness and competitiveness of battery electric and hybrid alternative technologies for the traction system. In response, the Spanish government aims to install 100 to 150 HRSs nationwide by 2030 amongst other objectives for the hydrogen infrastructure. This paper introduces a Spain-wide, RES-clustered design and assessment of the renewable HRS network that jointly (i) estimates light- and heavy-duty FCEV demand anchored to national targets, grounded forecasts and empirical refuelling patterns (ii) optimally decides per each RES zone the distribution and capacity using two station scales (daily hydrogen throughput capacity) and (iii) analyzes the on-grid and off-grid HRS design requirements and techno-economic results of their deployment. The clustering of the Spanish territory is performed to discretize the simulation and analysis geographically, and the clusters are decided based on RES potential. The main core of the techno-economic modelling and simulation is carried out via a coordinated HDRSAM–HOMER workflow that improves on common approaches presented in the literature, where only one of these tools is considered. The results reveal that off-grid stations have a Levelized Cost of Hydrogen (LCOH) ranging from 14.79 $/kg to 24.13 $/kg, with larger stations benefiting from modest economies of scale. By contrast, on-grid HRSs present lower LCOH (9.56–16.71 $/kg) and a significantly lower energy surplus than that in off-grid topologies. Consequently, grid connectivity proves advantageous in regards to energy utilization and cost. These findings provide a robust foundation for scaling a renewable HRS network rollout in Spain, underscoring the role of on-grid solutions in accelerating FCEV adoption and enhancing sustainable mobility while keeping RES input necessary to optimize costs and reduce emissions., This research has been developed within the CSIC Interdisciplinary Thematic Platform (PTI+) Transición Energética Sostenible+ (PTI-TRANSENER+)[TRE2103000] as part of the CSIC program for the Spanish Recovery, Transformation and Resilience Plan funded by the Recovery and Resilience Facility of the European Union, established by the Regulation (EU) 2020/2094; Project MASHED (TED2021-129927B-I00) funded by MCIN/ AEI /10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”; and the Project MAFALDA (PID2021-126001OB-C31) funded by MCIN/ AEI /10.13039/501100011033 and by “ERDF A way of making Europe ”; Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI /10.13039/501100011033/FEDER, UE., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Contribución de congreso (ConferenceOutput)

Detección y localización de fugas en redes de distribución de agua potable, Leak detection and localization in water distribution networks

  • Blesa, Joaquim
Trabajo presentado en las XLVI Jornadas de Automática, celebradas en Cartagena (España), del 3 al 5 de septiembre de 2025, Las redes de agua potable enfrentan graves pérdidas por fugas (hasta 30\%), agravando la escasez hídrica mundial. En Europa, se estima que el 25\% del agua tratada se pierde produciendo un coste económico y medioambiental importante. La UE y España impulsan soluciones mediante directivas y fondos para digitalización con el objetivo de minimizar estas pérdidas. Este trabajo aborda el problema de la detección y localización de fugas en redes de distribución de agua. En primer lugar, se presenta una revisión de los principales métodos existentes en la literatura técnica. Posteriormente, se presentan dos metodologías específicas: un método para la detección de fugas y otro para su localización precisa. Finalmente, estas propuestas metodológicas son implementadas y evaluadas mediante su aplicación a dos casos de estudio., Este trabajo forma parte del proyecto PID2023-148840OB-I00 (SEAMLESS), financiado por MCIN/ AEI., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Optimal Regional Water Network Management in Drought Conditions: the Catalonia Case Study

  • Vázquez-Vázquez, Abner
  • Segovia, Pablo
  • Puig, Vicenç
3rd IFAC Workshop on Control Methods for Water Resource Systems CMWRS 2025 Douai, France, November 6-7, 2025, In the present work, Model Predictive Control (MPC) is applied to the regional water network of Catalonia, which consists of several dams in the Llobregat and Ter rivers, a desalination plant and a treated water recirculation system. The objective is to evaluate the impact of these two latter alternative sources to guarantee water supply for prolonged drought periods. The MPC formulation minimizes the multi-objective cost function, considering specific operational objectives, such as flow controls, tank levels and ecological flows, weighted according to their relative importance. Two cases are analyzed: with and without alternative sources. The simulation shows an improvement by using alternative sources, significantly delaying water depletion and allowing for greater network resilience in with limited supply., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Adaptive scenario-based predictive control: a set membership learning approach

  • Meza Camargo, Erick Manuel
  • Del Duca, Alessandro
  • Ruiz, Fredy
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
This paper explores using the Set Membership estimation approach within a scenario-based Model Predictive Control framework to robustly control unknown constrained systems. By assuming unknown-but-bounded noises, a set of models compatible with available data is derived at each time step. Then, a stochastic predictive control scheme regulates the system, ensuring probabilistic operational constraint satisfaction. The combined strategy allows to iteratively refine the set of feasible models that explain the system dynamics, enhancing closed-loop performance as the system operates, while its computational complexity remains lower than that of similar solutions, requiring solving a quadratic program with a set of linear constraints that grows linearly with the risk level, and a set of linear programs when new information allows to update the models set for adaptation. Simulations related to depth tracking of a diving system demonstrate the effective operation of the strategy, reducing uncertainty while satisfying constraints. An initial unconstrained phase is also considered to further improve the identification phase, yielding notable performance gains while satisfying all constraints., This work has been supported by the Italian Ministry of University and Research under grant “Learning-based Model Predictive Control by Exploration and Exploitation in Uncertain Environments” (PRIN PNRR 2022 fund, ID P2022EXP2W) and the Spanish project SEAMLESS: “Sustainable learningbased Management of Multi-resource Large-scale Systems” (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Model predictive control for personalized abiraterone administration in metastatic castrate-resistant prostate cancer

  • Pena-Campos, Johan
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Patiño Guevara, Diego
  • Caicedo Dorado, Alexander
© 2025 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., Metastatic castrate-resistant prostate cancer (mCRPC) presents significant therapeutic difficulties. This study develops and evaluates a Model Predictive Control (MPC) framework for the personalized administration of Abiraterone in mCRPC. The proposed MPC strategy dynamically optimizes drug dosage over a finite prediction horizon, utilizing a well-established mathematical model of prostate cancer cell populations from the literature (androgen-dependent T+, testosterone-producing TP, and androgen-independent T-). This adaptive approach is designed to delay disease progression and manage tumor burden more effectively than static or pre-optimized schedules. Simulation results obtained in this paper demonstrate the effectiveness of the proposed controller. Specifically, the MPC approach shows potential in prolonging progression metrics, such as the time until resistant T- cells prevail, and in reducing cumulative drug exposure through adaptive dosing, thereby supporting the objective of managing cancer as a chronic condition., The work of this paper is supported by the following funds: the Condonable Credit Doctoral Support Grant from the Academic Vice Rector’s Office, the Scholarship Loan Program Grant from Colfuturo, the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE and the project Artificial intelligence modeling for early detection of pancreatic and stomach cancer under the ID 20346., Peer Reviewed, 3 - Salut i Benestar
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Nonlinear dynamic modeling and control of ethanol steam reforming for hydrogen production

  • Arcilla-Osorio, Mateo
  • Destro, Francesco
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Llorca, Jordi
  • Braatz, Richard D.
Copyright © 2025. The Authors. This is an open access article under the CC BY-NC-ND license., Hydrogen is a key facilitator of smart energy systems, with applications in power generation, transportation, and industry. This article presents a nonlinear dynamic model of ethanol steam reforming (ESR) within an staged separation membrane reactor (SSMR) and explores its application in real-time control. A proportional-integral-derivative (PID) controller is implemented to regulate hydrogen output by adjusting ethanol feed to match a step-wise demand profile. Simulation results confirm the model applicability and control effectiveness. This study demonstrates the potential of advanced modeling for hydrogen-based energy applications, paving the way for future research in advanced control design and broader energy systems integration., This work was supported in part by the Ag`encia de Gesti´o d’Ajuts Universitaris i de Recerca (AGAUR), Generalitat de Catalunya under Grant FI SDUR 2023 and the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Calibration of friction and roughness in an urban water distribution network using an LSTM neural network-based framework

  • Gómez Coronel, Leonardo
  • Torres Ortiz, Flor Lizeth
  • Santos Ruiz, Ildeberto de los
  • Blesa Izquierdo, Joaquim|||0000-0002-5626-3753
  • Ramírez Chavarría, Roberto
© 2025 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., This paper proposes a method to estimate both the friction factor and the roughness coefficient of the pipes in a water distribution network (WDN). Pressure head data is used for the training of a Long-Short Term Memory (LSTM) neural network to predict the pressure head at all of the other nodes in the network, as well as the flow rate through the pipes. When all the data is predicted, the estimated pressure head drops, and the flow rate for the entire water distribution network is used to compute both the friction factor and the flow rate. The proposed method is tested using the well-known Hanoi DMA (District Metered Area) as a case study. Satisfactory results are obtained when comparing the predicted friction factor and roughness coefficient against the expected values., The authors gratefully acknowledge the financial support provided by the Tecnologico Nacional de M ´ exico and SECIHTI (Mexico). The authors also ácknowledge the support provided by SECTEI-CDMX through project eSAST, number 1564c23, DGAPA-UNAM through project IT1000724, and by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Neural network-based leak localization in water distribution networks using the gravity center of pressure measurements

  • Gómez Coronel, Leonardo
  • Blesa Izquierdo, Joaquim|||0000-0002-5626-3753
  • Santos Ruiz, Ildeberto
  • López Estrada, Francisco Ronay
  • Puig Cayuela, Vicenç|||0000-0002-6364-6429
A novel methodology for leak diagnosis in urban water distribution systems (WDS) is proposed. Small leaks are simulated using a well-calibrated EPANET model of the WDS. Considering only the known topology of the WDS, and pressure head values recorded at some nodes, the center of gravity of pressure is computed. Under nominal (leak-free) operation the position of the center of gravity varies predictably, but leaks cause variations on its position. Sensor-measurements with a duration of 24 h are used to compute residual coordinates from leak-free operation and used to train a LSTM neural network implemented in MATLAB for leak classification. Results are presented for the leak localization task considering two levels of resolution: identifying the general sector and pinpointing the specific node where the leak occurs. Tests are performed on a benchmark and real-world WDS obtaining a good performance with simulated data under steady-state and variable demand conditions. The impact of measurement noise is addressed by including the measured outflow from the reservoir as a third dimension to the training data., This work was developed within the framework of RICCA “Red Internacional de Control y Cómputo Aplicados”. Thanks to CONAHCYT and Tecnológico Nacional de México for the funding granted for this research through project 20,212.24-P. We would also like to thank the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE for supporting this research., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Large-scale EV charging coordination: a detailed exploration of mean-field and reinforcement learning approaches

  • Caballé Benavent, Josep
  • Segovia Castillo, Pablo|||0000-0003-3593-907X
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Quijano Silva, Nicanor
The rapid proliferation of electric vehicles (EVs) presents significant challenges to power-grid stability, especially when thousands of cars charge simultaneously without coordination. This paper investigates two complementary families of scalable control schemes-(i) Mean-Field Control (Mean-Field Games and Mean-Field-Type Games), and (ii) model-free Reinforcement Learning (classical Reinforcement Learning and Deep Reinforcement Learn ing)-that capture stochastic arrivals, nodal capacity limits, ambient-temperature effects and battery degradation. Analytical mean-field-based formulations yield decentralized charging policies that depend only on population statistics and enjoy ¿-Nash optimality as fleet size grows, while reinforcement-learning-based agents learn di rectly from interaction histories and cope naturally with partial observability and non-stationary price signals. A common set of nine key performance indicators is applied to two benchmark scenarios: a one-night, three-node grid and a seven-day, heterogeneous 1 000-EV testbed built on real distribution-network data. Results show that Mean-Field-Type Games minimizes unmet energy (=1 %) and queueing delays, Deep Reinforcement Learning max imizes average final State-of-Charge (˜81 %) under volatile tariffs, and classical Reinforcement Learning provides the most interpretable albeit least efficient baseline. These quantified trade-offs clarify when model-based equi librium methods suffice and when adaptive, data-driven controllers become indispensable, providing actionable guidance for large-scale, battery-health-aware EV-charging deployments., The work of Josep Caballé is supported by CSIC, JAE Intro Programme. The work of Pablo Segovia and Carlos Ocampo-Martinez is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN / AEI / 10.13039 / 501100011033 / FEDER, UE. The work of Nicanor Quijano is partially supported by Asignación para la CTeI - SGR and Minciencias, Colombia, under project BPIN 2021000100499., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Interval luenberger-like observer design for fault detection and state estimation in one-sided Lipschitz quadratically inner-bounded systems

  • Arango Restrepo, Juan Pablo
  • Etienne, Lucien
  • Duviella, Eric
  • Langueh, Kokou
  • Segovia Castillo, Pablo|||0000-0003-3593-907X
  • Puig Cayuela, Vicenç|||0000-0002-6364-6429
© 2025 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., This paper presents the design of an interval observer (IO) for systems characterized by nonlinearities that are assumed to fulfill one-sided Lipschitz quadratically innerbounded (OSL-QIB) conditions. This novel approach extends conventional design of IO for Lipschitz systems to OSL-QIB systems with the aim of improving observer design and convergence. Additionally, this new IO design is applied to fault detection, and state estimation which is crucial to guarantee normal process operation. A bioreactor for biomass production and a FHN system are considered as case studies to test the performance of the approach for fault detection and state estimation purposes, respectively., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Neural network pruning using nonlinear oblique subspace projections

  • Pena-Campos, Johan
  • Caicedo Dorado, Alexander
  • Patiño Guevara, Diego
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
© 2026 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., The increasing complexity of deep neural networks often leads to overparameterization, limiting their deployment in resource-constrained environments. This paper introduces a novel pruning methodology based on Nonlinear Oblique Subspace Projections (NObSP), an algorithm originally developed to enhance the interpretability of kernel regression models. In this work, NObSP is extended to neural networks to estimate the contribution of individual neurons and entire layers to the network's output. This is achieved by quantifying the nonlinear transformations performed by the network on its inputs and internal representations. By computing the magnitude of these variables, neurons contributing minimally to the network's output are identified. Specifically, neurons with a nonlinear effect magnitude falling below a defined threshold are pruned, effectively removing them from the network. Consequently, NObSP provides a data-driven, interpretability-based approach to neural network pruning, leading to simplified and efficient models with minor performance impact. This method offers a clear framework for understanding which components of a neural network are critical for its function, facilitating the development of more compact and interpretable deep learning models. Finally, the efficacy of NObSP for pruning is demonstrated in CIFAR-10 and MNIST classification problems, where it is compared to traditional pruning algorithms based on the Ln norm., The work of this paper is supported by the following funds: the Condonable Credit Doctoral Support Grant from the Academic Vice Rector’s Office, the Scholarship Loan Program Grant from Colfuturo, the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OBI00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE and the project PID2023-146261OB-I00, funded by Ministry of Science and Innovation of Spain., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Online current signature analysis approach for cavitation assessment in hydraulic pumps

  • Cruz Rangel, David Antonio|||0000-0003-3707-5279
  • Perurena Landache, Naroa
  • De-la-Fuente Azpeitia, Naiara
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Díaz Rozo, Javier
Non-intrusive load monitoring provides an efficient method for assessing the operational state of various equipment, including pumps, by identifying abnormal behaviour such as cavitation. However, Traditional electromechanical monitoring techniques often rely on fixed operating speeds or require expensive sensors, such as accelerometers, limiting their applicability in industrial environments where machines operate under constant and variable speed conditions. This study proposes a novel non-intrusive condition monitoring methodology capable of detecting cavitation under stable and fuctuating speed conditions. The approach is based on real-time electrical current signature feature extraction combined with a machine learning framework to assess the presence and severity of cavitation in mechanical components. The methodology has been validated through a practical case study involving heating, ventilation, and air conditioning (HVAC) water pumps. The results demonstrate the effectiveness of this approach in enhancing predictive maintenance strategies by enabling early cavitation detection, thereby reducing both energy inefficiencies and potential mechanical damage., This work has been supported by the Doctorats In- dustrials program from the Catalonia Government (Ref. 2022 DI 00093) and the technology transfer project IN- SPIRED (Ref. C-12727-UPC). Also, it was partially sup- ported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-20232016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the PID2023-148840OB-I00 (SEAMLESS) projects., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Optimal regional water network management in drought conditions: the Catalonia case study

  • Vázquez Vázquez, Abner
  • Segovia Castillo, Pablo|||0000-0003-3593-907X
  • Puig Cayuela, Vicenç|||0000-0002-6364-6429
Copyright © 2025 The Authors. This is an open access article under the CC BY-NC-ND license, In the present work, Model Predictive Control (MPC) is applied to the regional water network of Catalonia, which consists of several dams in the Llobregat and Ter rivers, a desalination plant and a treated water recirculation system. The objective is to evaluate the impact of these two latter alternative sources to guarantee water supply for prolonged drought periods. The MPC formulation minimizes the multi-objective cost function, considering specific operational objectives, such as flow controls, tank levels and ecological flows, weighted according to their relative importance. Two cases are analyzed: with and without alternative sources. The simulation shows an improvement by using alternative sources, significantly delaying water depletion and allowing for greater network resilience in with limited supply., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

An observer-based tracking controller for one sided Lipschitz – quadratically inner bounded perturbed nonlinear systems

  • Arango Restrepo, Juan Pablo
  • Etienne, Lucien
  • Duviella, Eric
  • Langueh, Kokou
  • Segovia Castillo, Pablo|||0000-0003-3593-907X
  • Puig Cayuela, Vicenç|||0000-0002-6364-6429
This is an original manuscript of an article published by Taylor & Francis in International journal of systems science on 18 Jan 2026, available at: https://doi.org/10.1080/00207721.2025.2601783., This paper presents the design of an observer-based tracking controller for one-sided Lipschitz (OSL) Quadratically Inner Bounded (QIB) Nonlinear Systems considering perturbations in the output and reference tracking. Initially, the observer and controller synthesis procedures are formulated individually in the linear matrix inequality (LMI) framework. Then, a separation principle is established where it is proved that the separately designing observer and controller (guaranteeing error stability), when coupled, continue to converge. After having the convergence conditions of observer-based control (OBC), we proceed to show a numerical example to corroborate the theoretical developments in practice., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A nonlinear state feedback tracking controller for open channel irrigation systems

  • Arango Restrepo, Juan Pablo
  • Duviella, Eric
  • Etienne, Lucien
  • Langueh, Kokou
  • Segovia Castillo, Pablo|||0000-0003-3593-907X
  • Puig Cayuela, Vicenç|||0000-0002-6364-6429
The majority of human activities requires water, which is mostly transported and supplied using open channel irrigation systems (OCIS). Their optimal operation requires an accurate control to reject disturbances that can arise under day-to-day operation (e.g., rainfall, leaks, and theft) and thus guarantee user satisfaction and safety. This paper proposes the design of a control strategy, we use part of the OSL-QIB methodology previously introduced by Arango et al. (2024) with a linear parameter varying term (LPV) to handle non-linear affine terms more easily than other commonly used methods, such as Feedback Linearization. From a phenomenological point of view, the considered model to construct the nonlinear control strategy is based on material and energy balances, whose advantage is to avoid delays (very common in OCIS modeling) and capture important phenomena that occurs in the process. The proposed controller is tested in the first two pools of the Corning channel, obtaining satisfactory results (disturbance rejection and tracking)., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Cloud computing integration for leak diagnosis using meta-heuristic methods for water distribution networks

  • Sánchez Mejía, Luis Adan
  • Gómez Coronel, Leonardo
  • Blesa Izquierdo, Joaquim|||0000-0002-5626-3753
  • Torres Ortiz, Flor Lizeth
  • Santos Ruiz, Ildeberto de los
© 2025 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., We present the implementation of a CloudComputing-based system for leak diagnosis in a water distribution network. Pressure head/flow rate measured in field is stored and processed in a virtual machine to provide a leak diagnosis in two stages: 1) first the leak is detected by comparing the current operating conditions and expected nominal operating conditions obtained from a previously adjusted simulation model, and 2) the leak exact location and its magnitude are determined using a meta-heuristic method. The performance of the proposed system is implemented for an experimental hydraulic system at a laboratory scale. Results demonstrate a good accuracy in the leak diagnosis metrics at a reduced economic and computational cost, demonstrating the potential results from the implementation of a similar system in a large-scale water distribution network., The authors gratefully acknowledge the financial support provided by the Tecnológico Nacional de México and SECIHTI (Mexico). The authors also acknowledge the support provided by SECTEI-CDMX through project eSAST, number 1564c23, DGAPA-UNAM through project IT1000724, and by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A benchmark simulator for advanced control of ethanol steam reforming

  • Arcila Osorio, Mateo|||0009-0009-9719-4721
  • Destro, Francesco
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Llorca Piqué, Jordi|||0000-0002-7447-9582
  • Braatz, Richard D.
© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/)., This article presents a nonlinear dynamic simulator of an ethanol steam reforming process designed for pure hydrogen production for fuel-cell applications. The process consists of two stages: alcohol reforming and hydrogen separation within a staged-separation membrane reactor. The simulator is intended to serve as a benchmark for developing and testing advanced control, estimation, and optimization strategies. The simulator is extensively validated with experimental data. Detailed descriptions of the experimental setup, key modeling assumptions, and reference operating conditions are provided. Furthermore, critical control challenges that must be addressed to achieve optimal process operation are identified, along with quantitative metrics for evaluating the performance of control strategies implemented by users. A set of case studies illustrate the implementation of a proportional–integral–derivative (PID) controller in the benchmark simulator for tracking a hydrogen flow set-point and effectively rejecting disturbances. The benchmark simulator is developed in MATLAB® and is available in an online repository., This work was supported in part by the Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR), Generalitat de Catalunya, Spain under Grant FI SDUR 2023 FISDU 00160 and the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Mixed logical dynamical modelling of renewable hydrogen refuelling stations for the design of optimization-based operational schemes

  • Cardona Rubio, Pol|||0000-0001-7504-2916
  • Valiño Garcia, Luis
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Serra, Maria|||0000-0002-9885-8093
© 2025 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/)., This paper proposes a Mixed Logical Dynamical (MLD) model for a real-world Hydrogen Refuelling Station (HRS) currently under development, incorporating on-site production, multi-pressure hydrogen storage, and discrete event-triggered refuelling processes. By modelling the HRS using the expanded linear state-space MLD formulation with linear inequality constraints, the model accommodates hydrogen flows, pressure thresholds, and discontinuous behaviour within this unified framework suitable for automatic control/decision-making purposes. To illustrate the usefulness of the proposed modelling approach, a preliminary Hybrid Model Predictive Control (HMPC) scheme and a Constraint Satisfaction Problem (CSP) formulation are proposed, leveraging the MLD structure for optimization-based control and feasibility validation in the face of unpredicted Fuel-Cell Electric Vehicle (FCEV) arrivals. The case of study simulation results highlight how the MLD model addresses the logical and event-triggered behaviours and constraints commonly neglected by aggregated models reported in scheduling-based approaches present in the gross body of literature concerning HRS operation. The implemented HMPC strategy and the CSP statement primarily demonstrate the model’s practical utility. These approaches also hint at their potential for developing advanced operational strategies—ranging from stochastic or multi-level control schemes to distributed architectures—and for deeper sizing analyses or performance assessments of real-world HRSs. Consequently, the proposed modelling approach provides a robust foundation for innovation in hydrogen infrastructure management, bridging essential gaps in the literature by integrating discrete logic decisions, multi-tank refuelling topologies, and online optimization under real operating constraints., This research has been developed within the CSIC Interdisciplinary Thematic Platform (PTI+) Transición Energética Sostenible+ (PTI-TRANSENER+)[TRE2103000] as part of the CSIC program for the Spanish Recovery, Transformation and Resilience Plan funded by the Recovery and Resilience Facility of the European Union, established by the Regulation (EU) 2020/2094; Project MASHED (TED2021-129927B-I00) funded by MCIN/AEI/10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”; and the Project MAFALDA (PID2021- 126001OB-C31) funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe”; Spanish project SEAMLESS: Sustainable Learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Novel model-based framework for multi-fault analysis in ball bearings

  • Arenas Angulo, Vanessa Inés
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Díaz Rozo, Javier
© 2025 The Authors. This is an open access article under the CC BY-NC-ND, This paper presents a novel framework integrating fault models into a semi-physical, non-linear model calibrated with real sensor data to enhance fault analysis in ball- bearings. Unlike conventional methods that separate physics-based modeling and data-driven diagnostics, this approach merges both to improve fault assessment and predictive maintenance. The model accounts for degradation mechanisms such as wear, crack propagation, and creep, capturing their effects on mass, stiffness, and damping. This framework offers a reliable tool for bearing health monitoring, facilitating early fault detection and reducing industrial operational risks., This work has been supported by the technology transfer project INSPIRED (Ref. C-12727-UPC). Also, it was partially supported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-2023-2016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the SEAMLESS project (Ref. PID2023-148840OB-I00)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Advanced fault detection in steam turbines using Welch’s method and probabilistic modeling, Detección avanzada de fallas en turbinas de vapor mediante el método de Welch y modelado probabilístico

  • Puerto Santana, Cristian|||0000-0002-7380-5198
  • Lizarraga Odriozola, Nagore Lei
  • Díaz Rozo, Javier
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
Rotodynamic systems using steam turbines are essential for power generation, requiring reliable operation under continuous use. Mechanical vibrations are the major cause of failures, often linked to component degradation and structural resonances that reduce lifespan. Existing fault detection methods use signal processing, probabilistic models, and finite element analysis but struggle with scalability, real-time applicability, and integration of structural and operational data. This paper introduces a data-driven approach combining accelerometers, signal processing, and probabilistic modeling to detect anomalies. Validated on a real power system, the methodology effectively identifies resonance-related faults in rotodynamic vibrations over long-term monitoring., This work has been supported by the Doctorats Industrials program from the Catalonia Government (Ref. 2022DI 00093) and the technology transfer project INSPIRED (Ref.C-12727-UPC). Also, it was partially supported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-20232016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the PID2023-148840OB-I00 (SEAMLESS) projects., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

An Integrated novel random decrement technique and system identification framework for modal features extraction in structural health monitoring

  • Puerto Santana, Cristian|||0000-0002-7380-5198
  • Diaz Rozo, Javier
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
Structural health monitoring (SHM) techniques are used for monitoring the condition of structures, including bridges, buildings, dams, among others, through data acquired from embedded sensors. An integral aspect of SHM involves feature extraction, wherein algorithms analyze raw sensor data to generate explainable features regarding the health state of structures. The random decrement technique (RDT) is a notable signal processing method used to estimate the free response of mechanical structures, crucial for extracting frequency features related to structural dynamics. However, studies have shown that RDT encounters performance issues in cases of nonlinear damping, significant signal noise, and overlapping vibration modes. In a previous study, a data-driven classifier for transient and stationary regimes in multivariate dynamic systems was introduced, employing differential geometry tools to analyze stochastic properties. In this paper, the authors applied this classifier to develop a novel frequency-feature extraction method for SHM. This method incorporates a modified triggering condition for RDT and a recent system identification technique, aiming to improve frequency-feature estimation, especially in segment selection and modal parameter estimation of free responses in non-linear structures. Validation of the proposed approach employed mechanical vibration data from four accelerometer sensors on the “Andoian DF 12” bridge in Guipuzkoa, Spain, covering a one-year period and providing comprehensive insights into structural dynamics across different seasons. Comparative analysis against traditional RDT and various system identification techniques demonstrated that the proposed methodology improve the estimation of structural free responses by reducing noise and requiring fewer data compared to traditional RDTs. Moreover, the estimated modes aligned with the spectral content of mechanical sensors, indicating suitability for online applications with minimal computational costs and accurate modal parameter estimation for SHM in structures., This work has been supported by the Doctorats Industrials program from the Catalonia Government (Ref. [2022] DI 00093) and the technology transfer project INSPIRED (Ref. C-12727-UPC). Also, it was partially supported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-20232016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the PID2023-148840OB-I00 (SEAMLESS) projects., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Partitioning and distributed predictive control of UAV swarms

  • Delansnay, Gilles|||0000-0003-3007-8493
  • Ocampo-Martínez, Carlos|||0000-0001-9251-6044
  • Dewasme, Laurent|||0000-0002-1688-167X
  • Vande Wouwer, Alain|||0000-0001-7022-6126
Coordinating large swarms of unmanned aerial vehicles (UAVs) is a complex task due to high dimensionality and real-time computational demand. Effective formation control requires scalable strategies to maintain desired configurations while ensuring robustness to disturbances. This paper proposes a graph-based partitioning approach that divides the swarm into smaller, manageable subgroups, each of which is coordinated through an inter-partition communication network. This algorithm enables decentralized control, reducing computational complexity and enhancing scalability. This framework is integrated with a distributed control strategy, e.g., tube model predictive control, to ensure robust formation under uncertainties. Realistic simulations demonstrate the effectiveness of the proposed approach for UAV swarms., The work of C. Ocampo-Martinez is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE ., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

An observer-based tracking controller for one sided Lipschitz – quadratically inner bounded perturbed nonlinear systems

  • Arango Restrepo, Juan Pablo
  • Etienne, Lucien
  • Duviella, Eric
  • Langueh, Kokou
  • Segovia Castillo, Pablo
  • Puig Cayuela, Vicenç
This is an original manuscript of an article published by Taylor & Francis in International journal of systems science on 18 Jan 2026, available at: https://doi.org/10.1080/00207721.2025.2601783., This paper presents the design of an observer-based tracking controller for one-sided Lipschitz (OSL) Quadratically Inner Bounded (QIB) Nonlinear Systems considering perturbations in the output and reference tracking. Initially, the observer and controller synthesis procedures are formulated individually in the linear matrix inequality (LMI) framework. Then, a separation principle is established where it is proved that the separately designing observer and controller (guaranteeing error stability), when coupled, continue to converge. After having the convergence conditions of observer-based control (OBC), we proceed to show a numerical example to corroborate the theoretical developments in practice., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A nonlinear state feedback tracking controller for open channel irrigation systems

  • Arango Restrepo, Juan Pablo
  • Duviella, Eric
  • Etienne, Lucien
  • Langueh, Kokou
  • Segovia Castillo, Pablo
  • Puig Cayuela, Vicenç
The majority of human activities requires water, which is mostly transported and supplied using open channel irrigation systems (OCIS). Their optimal operation requires an accurate control to reject disturbances that can arise under day-to-day operation (e.g., rainfall, leaks, and theft) and thus guarantee user satisfaction and safety. This paper proposes the design of a control strategy, we use part of the OSL-QIB methodology previously introduced by Arango et al. (2024) with a linear parameter varying term (LPV) to handle non-linear affine terms more easily than other commonly used methods, such as Feedback Linearization. From a phenomenological point of view, the considered model to construct the nonlinear control strategy is based on material and energy balances, whose advantage is to avoid delays (very common in OCIS modeling) and capture important phenomena that occurs in the process. The proposed controller is tested in the first two pools of the Corning channel, obtaining satisfactory results (disturbance rejection and tracking)., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Advanced fault detection in steam turbines using Welch’s method and probabilistic modeling, Detección avanzada de fallas en turbinas de vapor mediante el método de Welch y modelado probabilístico

  • Puerto Santana, Cristian
  • Lizarraga Odriozola, Nagore Lei
  • Díaz Rozo, Javier
  • Ocampo-Martínez, Carlos
Rotodynamic systems using steam turbines are essential for power generation, requiring reliable operation under continuous use. Mechanical vibrations are the major cause of failures, often linked to component degradation and structural resonances that reduce lifespan. Existing fault detection methods use signal processing, probabilistic models, and finite element analysis but struggle with scalability, real-time applicability, and integration of structural and operational data. This paper introduces a data-driven approach combining accelerometers, signal processing, and probabilistic modeling to detect anomalies. Validated on a real power system, the methodology effectively identifies resonance-related faults in rotodynamic vibrations over long-term monitoring., This work has been supported by the Doctorats Industrials program from the Catalonia Government (Ref. 2022DI 00093) and the technology transfer project INSPIRED (Ref.C-12727-UPC). Also, it was partially supported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-20232016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the PID2023-148840OB-I00 (SEAMLESS) projects., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Neural network-based leak localization in water distribution networks using the gravity center of pressure measurements

  • Gómez Coronel, Leonardo
  • Blesa Izquierdo, Joaquim
  • Santos Ruiz, Ildeberto
  • López Estrada, Francisco Ronay
  • Puig Cayuela, Vicenç
A novel methodology for leak diagnosis in urban water distribution systems (WDS) is proposed. Small leaks are simulated using a well-calibrated EPANET model of the WDS. Considering only the known topology of the WDS, and pressure head values recorded at some nodes, the center of gravity of pressure is computed. Under nominal (leak-free) operation the position of the center of gravity varies predictably, but leaks cause variations on its position. Sensor-measurements with a duration of 24 h are used to compute residual coordinates from leak-free operation and used to train a LSTM neural network implemented in MATLAB for leak classification. Results are presented for the leak localization task considering two levels of resolution: identifying the general sector and pinpointing the specific node where the leak occurs. Tests are performed on a benchmark and real-world WDS obtaining a good performance with simulated data under steady-state and variable demand conditions. The impact of measurement noise is addressed by including the measured outflow from the reservoir as a third dimension to the training data., This work was developed within the framework of RICCA “Red Internacional de Control y Cómputo Aplicados”. Thanks to CONAHCYT and Tecnológico Nacional de México for the funding granted for this research through project 20,212.24-P. We would also like to thank the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE for supporting this research., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Novel model-based framework for multi-fault analysis in ball bearings

  • Arenas Angulo, Vanessa Inés
  • Ocampo-Martínez, Carlos
  • Díaz Rozo, Javier
© 2025 The Authors. This is an open access article under the CC BY-NC-ND, This paper presents a novel framework integrating fault models into a semi-physical, non-linear model calibrated with real sensor data to enhance fault analysis in ball- bearings. Unlike conventional methods that separate physics-based modeling and data-driven diagnostics, this approach merges both to improve fault assessment and predictive maintenance. The model accounts for degradation mechanisms such as wear, crack propagation, and creep, capturing their effects on mass, stiffness, and damping. This framework offers a reliable tool for bearing health monitoring, facilitating early fault detection and reducing industrial operational risks., This work has been supported by the technology transfer project INSPIRED (Ref. C-12727-UPC). Also, it was partially supported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-2023-2016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the SEAMLESS project (Ref. PID2023-148840OB-I00)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A benchmark simulator for advanced control of ethanol steam reforming

  • Arcila Osorio, Mateo
  • Destro, Francesco
  • Ocampo-Martínez, Carlos
  • Llorca Piqué, Jordi
  • Braatz, Richard D.
© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/)., This article presents a nonlinear dynamic simulator of an ethanol steam reforming process designed for pure hydrogen production for fuel-cell applications. The process consists of two stages: alcohol reforming and hydrogen separation within a staged-separation membrane reactor. The simulator is intended to serve as a benchmark for developing and testing advanced control, estimation, and optimization strategies. The simulator is extensively validated with experimental data. Detailed descriptions of the experimental setup, key modeling assumptions, and reference operating conditions are provided. Furthermore, critical control challenges that must be addressed to achieve optimal process operation are identified, along with quantitative metrics for evaluating the performance of control strategies implemented by users. A set of case studies illustrate the implementation of a proportional–integral–derivative (PID) controller in the benchmark simulator for tracking a hydrogen flow set-point and effectively rejecting disturbances. The benchmark simulator is developed in MATLAB® and is available in an online repository., This work was supported in part by the Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR), Generalitat de Catalunya, Spain under Grant FI SDUR 2023 FISDU 00160 and the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Mixed logical dynamical modelling of renewable hydrogen refuelling stations for the design of optimization-based operational schemes

  • Cardona Rubio, Pol
  • Valiño Garcia, Luis
  • Ocampo-Martínez, Carlos
  • Serra, Maria
© 2025 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/)., This paper proposes a Mixed Logical Dynamical (MLD) model for a real-world Hydrogen Refuelling Station (HRS) currently under development, incorporating on-site production, multi-pressure hydrogen storage, and discrete event-triggered refuelling processes. By modelling the HRS using the expanded linear state-space MLD formulation with linear inequality constraints, the model accommodates hydrogen flows, pressure thresholds, and discontinuous behaviour within this unified framework suitable for automatic control/decision-making purposes. To illustrate the usefulness of the proposed modelling approach, a preliminary Hybrid Model Predictive Control (HMPC) scheme and a Constraint Satisfaction Problem (CSP) formulation are proposed, leveraging the MLD structure for optimization-based control and feasibility validation in the face of unpredicted Fuel-Cell Electric Vehicle (FCEV) arrivals. The case of study simulation results highlight how the MLD model addresses the logical and event-triggered behaviours and constraints commonly neglected by aggregated models reported in scheduling-based approaches present in the gross body of literature concerning HRS operation. The implemented HMPC strategy and the CSP statement primarily demonstrate the model’s practical utility. These approaches also hint at their potential for developing advanced operational strategies—ranging from stochastic or multi-level control schemes to distributed architectures—and for deeper sizing analyses or performance assessments of real-world HRSs. Consequently, the proposed modelling approach provides a robust foundation for innovation in hydrogen infrastructure management, bridging essential gaps in the literature by integrating discrete logic decisions, multi-tank refuelling topologies, and online optimization under real operating constraints., This research has been developed within the CSIC Interdisciplinary Thematic Platform (PTI+) Transición Energética Sostenible+ (PTI-TRANSENER+)[TRE2103000] as part of the CSIC program for the Spanish Recovery, Transformation and Resilience Plan funded by the Recovery and Resilience Facility of the European Union, established by the Regulation (EU) 2020/2094; Project MASHED (TED2021-129927B-I00) funded by MCIN/AEI/10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”; and the Project MAFALDA (PID2021- 126001OB-C31) funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe”; Spanish project SEAMLESS: Sustainable Learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE, Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Nonlinear dynamic modeling and control of ethanol steam reforming for hydrogen production

  • Arcilla-Osorio, Mateo
  • Destro, Francesco
  • Ocampo-Martínez, Carlos
  • Llorca, Jordi
  • Braatz, Richard D.
Copyright © 2025. The Authors. This is an open access article under the CC BY-NC-ND license., Hydrogen is a key facilitator of smart energy systems, with applications in power generation, transportation, and industry. This article presents a nonlinear dynamic model of ethanol steam reforming (ESR) within an staged separation membrane reactor (SSMR) and explores its application in real-time control. A proportional-integral-derivative (PID) controller is implemented to regulate hydrogen output by adjusting ethanol feed to match a step-wise demand profile. Simulation results confirm the model applicability and control effectiveness. This study demonstrates the potential of advanced modeling for hydrogen-based energy applications, paving the way for future research in advanced control design and broader energy systems integration., This work was supported in part by the Ag`encia de Gesti´o d’Ajuts Universitaris i de Recerca (AGAUR), Generalitat de Catalunya under Grant FI SDUR 2023 and the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Online current signature analysis approach for cavitation assessment in hydraulic pumps

  • Cruz Rangel, David Antonio
  • Perurena Landache, Naroa
  • De-la-Fuente Azpeitia, Naiara
  • Ocampo-Martínez, Carlos
  • Díaz Rozo, Javier
Non-intrusive load monitoring provides an efficient method for assessing the operational state of various equipment, including pumps, by identifying abnormal behaviour such as cavitation. However, Traditional electromechanical monitoring techniques often rely on fixed operating speeds or require expensive sensors, such as accelerometers, limiting their applicability in industrial environments where machines operate under constant and variable speed conditions. This study proposes a novel non-intrusive condition monitoring methodology capable of detecting cavitation under stable and fuctuating speed conditions. The approach is based on real-time electrical current signature feature extraction combined with a machine learning framework to assess the presence and severity of cavitation in mechanical components. The methodology has been validated through a practical case study involving heating, ventilation, and air conditioning (HVAC) water pumps. The results demonstrate the effectiveness of this approach in enhancing predictive maintenance strategies by enabling early cavitation detection, thereby reducing both energy inefficiencies and potential mechanical damage., This work has been supported by the Doctorats In- dustrials program from the Catalonia Government (Ref. 2022 DI 00093) and the technology transfer project IN- SPIRED (Ref. C-12727-UPC). Also, it was partially sup- ported by the Spanish Centre for the Development of Industrial Technology (CDTI) through the MIG-20232016, INDUSTR-IA project and the Spanish Ministry of Science and Innovation through the PID2023-148840OB-I00 (SEAMLESS) projects., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Optimal regional water network management in drought conditions: the Catalonia case study

  • Vázquez Vázquez, Abner
  • Segovia Castillo, Pablo
  • Puig Cayuela, Vicenç
Copyright © 2025 The Authors. This is an open access article under the CC BY-NC-ND license, In the present work, Model Predictive Control (MPC) is applied to the regional water network of Catalonia, which consists of several dams in the Llobregat and Ter rivers, a desalination plant and a treated water recirculation system. The objective is to evaluate the impact of these two latter alternative sources to guarantee water supply for prolonged drought periods. The MPC formulation minimizes the multi-objective cost function, considering specific operational objectives, such as flow controls, tank levels and ecological flows, weighted according to their relative importance. Two cases are analyzed: with and without alternative sources. The simulation shows an improvement by using alternative sources, significantly delaying water depletion and allowing for greater network resilience in with limited supply., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Cloud computing integration for leak diagnosis using meta-heuristic methods for water distribution networks

  • Sánchez Mejía, Luis Adan
  • Gómez Coronel, Leonardo
  • Blesa Izquierdo, Joaquim
  • Torres Ortiz, Flor Lizeth
  • Santos Ruiz, Ildeberto de los
© 2025 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., We present the implementation of a CloudComputing-based system for leak diagnosis in a water distribution network. Pressure head/flow rate measured in field is stored and processed in a virtual machine to provide a leak diagnosis in two stages: 1) first the leak is detected by comparing the current operating conditions and expected nominal operating conditions obtained from a previously adjusted simulation model, and 2) the leak exact location and its magnitude are determined using a meta-heuristic method. The performance of the proposed system is implemented for an experimental hydraulic system at a laboratory scale. Results demonstrate a good accuracy in the leak diagnosis metrics at a reduced economic and computational cost, demonstrating the potential results from the implementation of a similar system in a large-scale water distribution network., The authors gratefully acknowledge the financial support provided by the Tecnológico Nacional de México and SECIHTI (Mexico). The authors also acknowledge the support provided by SECTEI-CDMX through project eSAST, number 1564c23, DGAPA-UNAM through project IT1000724, and by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Interval luenberger-like observer design for fault detection and state estimation in one-sided Lipschitz quadratically inner-bounded systems

  • Arango Restrepo, Juan Pablo
  • Etienne, Lucien
  • Duviella, Eric
  • Langueh, Kokou
  • Segovia Castillo, Pablo
  • Puig Cayuela, Vicenç
© 2025 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., This paper presents the design of an interval observer (IO) for systems characterized by nonlinearities that are assumed to fulfill one-sided Lipschitz quadratically innerbounded (OSL-QIB) conditions. This novel approach extends conventional design of IO for Lipschitz systems to OSL-QIB systems with the aim of improving observer design and convergence. Additionally, this new IO design is applied to fault detection, and state estimation which is crucial to guarantee normal process operation. A bioreactor for biomass production and a FHN system are considered as case studies to test the performance of the approach for fault detection and state estimation purposes, respectively., The work of this paper is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Calibration of friction and roughness in an urban water distribution network using an LSTM neural network-based framework

  • Gómez Coronel, Leonardo
  • Torres Ortiz, Flor Lizeth
  • Santos Ruiz, Ildeberto de los
  • Blesa Izquierdo, Joaquim
  • Ramírez Chavarría, Roberto
© 2025 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., This paper proposes a method to estimate both the friction factor and the roughness coefficient of the pipes in a water distribution network (WDN). Pressure head data is used for the training of a Long-Short Term Memory (LSTM) neural network to predict the pressure head at all of the other nodes in the network, as well as the flow rate through the pipes. When all the data is predicted, the estimated pressure head drops, and the flow rate for the entire water distribution network is used to compute both the friction factor and the flow rate. The proposed method is tested using the well-known Hanoi DMA (District Metered Area) as a case study. Satisfactory results are obtained when comparing the predicted friction factor and roughness coefficient against the expected values., The authors gratefully acknowledge the financial support provided by the Tecnologico Nacional de M ´ exico and SECIHTI (Mexico). The authors also ácknowledge the support provided by SECTEI-CDMX through project eSAST, number 1564c23, DGAPA-UNAM through project IT1000724, and by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Large-scale EV charging coordination: a detailed exploration of mean-field and reinforcement learning approaches

  • Caballé Benavent, Josep
  • Segovia Castillo, Pablo
  • Ocampo-Martínez, Carlos
  • Quijano Silva, Nicanor
The rapid proliferation of electric vehicles (EVs) presents significant challenges to power-grid stability, especially when thousands of cars charge simultaneously without coordination. This paper investigates two complementary families of scalable control schemes-(i) Mean-Field Control (Mean-Field Games and Mean-Field-Type Games), and (ii) model-free Reinforcement Learning (classical Reinforcement Learning and Deep Reinforcement Learn ing)-that capture stochastic arrivals, nodal capacity limits, ambient-temperature effects and battery degradation. Analytical mean-field-based formulations yield decentralized charging policies that depend only on population statistics and enjoy ¿-Nash optimality as fleet size grows, while reinforcement-learning-based agents learn di rectly from interaction histories and cope naturally with partial observability and non-stationary price signals. A common set of nine key performance indicators is applied to two benchmark scenarios: a one-night, three-node grid and a seven-day, heterogeneous 1 000-EV testbed built on real distribution-network data. Results show that Mean-Field-Type Games minimizes unmet energy (=1 %) and queueing delays, Deep Reinforcement Learning max imizes average final State-of-Charge (˜81 %) under volatile tariffs, and classical Reinforcement Learning provides the most interpretable albeit least efficient baseline. These quantified trade-offs clarify when model-based equi librium methods suffice and when adaptive, data-driven controllers become indispensable, providing actionable guidance for large-scale, battery-health-aware EV-charging deployments., The work of Josep Caballé is supported by CSIC, JAE Intro Programme. The work of Pablo Segovia and Carlos Ocampo-Martinez is partially supported by the Spanish project SEAMLESS: Sustainable learning-based Management of Multi-resource Large-scale Systems (ref. PID2023-148840OB-I00), funded by MCIN / AEI / 10.13039 / 501100011033 / FEDER, UE. The work of Nicanor Quijano is partially supported by Asignación para la CTeI - SGR and Minciencias, Colombia, under project BPIN 2021000100499., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Adaptive scenario-based predictive control: a set membership learning approach

  • Meza Camargo, Erick Manuel
  • Del Duca, Alessandro
  • Ruiz, Fredy
  • Ocampo-Martínez, Carlos
This paper explores using the Set Membership estimation approach within a scenario-based Model Predictive Control framework to robustly control unknown constrained systems. By assuming unknown-but-bounded noises, a set of models compatible with available data is derived at each time step. Then, a stochastic predictive control scheme regulates the system, ensuring probabilistic operational constraint satisfaction. The combined strategy allows to iteratively refine the set of feasible models that explain the system dynamics, enhancing closed-loop performance as the system operates, while its computational complexity remains lower than that of similar solutions, requiring solving a quadratic program with a set of linear constraints that grows linearly with the risk level, and a set of linear programs when new information allows to update the models set for adaptation. Simulations related to depth tracking of a diving system demonstrate the effective operation of the strategy, reducing uncertainty while satisfying constraints. An initial unconstrained phase is also considered to further improve the identification phase, yielding notable performance gains while satisfying all constraints., This work has been supported by the Italian Ministry of University and Research under grant “Learning-based Model Predictive Control by Exploration and Exploitation in Uncertain Environments” (PRIN PNRR 2022 fund, ID P2022EXP2W) and the Spanish project SEAMLESS: “Sustainable learningbased Management of Multi-resource Large-scale Systems” (ref. PID2023-148840OB-I00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Neural network pruning using nonlinear oblique subspace projections

  • Pena-Campos, Johan
  • Caicedo Dorado, Alexander
  • Patiño Guevara, Diego
  • Ocampo-Martínez, Carlos
© 2026 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., The increasing complexity of deep neural networks often leads to overparameterization, limiting their deployment in resource-constrained environments. This paper introduces a novel pruning methodology based on Nonlinear Oblique Subspace Projections (NObSP), an algorithm originally developed to enhance the interpretability of kernel regression models. In this work, NObSP is extended to neural networks to estimate the contribution of individual neurons and entire layers to the network's output. This is achieved by quantifying the nonlinear transformations performed by the network on its inputs and internal representations. By computing the magnitude of these variables, neurons contributing minimally to the network's output are identified. Specifically, neurons with a nonlinear effect magnitude falling below a defined threshold are pruned, effectively removing them from the network. Consequently, NObSP provides a data-driven, interpretability-based approach to neural network pruning, leading to simplified and efficient models with minor performance impact. This method offers a clear framework for understanding which components of a neural network are critical for its function, facilitating the development of more compact and interpretable deep learning models. Finally, the efficacy of NObSP for pruning is demonstrated in CIFAR-10 and MNIST classification problems, where it is compared to traditional pruning algorithms based on the Ln norm., The work of this paper is supported by the following funds: the Condonable Credit Doctoral Support Grant from the Academic Vice Rector’s Office, the Scholarship Loan Program Grant from Colfuturo, the Spanish project SEAMLESS: Sustainable learning-based Management of Multiresource Large-scale Systems (ref. PID2023-148840OBI00), funded by MCIN/AEI/10.13039/501100011033/FEDER, UE and the project PID2023-146261OB-I00, funded by Ministry of Science and Innovation of Spain., Peer Reviewed, Postprint (author's final draft)