Pub Date : 2026-03-01Epub Date: 2026-02-27DOI: 10.1016/j.ifacsc.2026.100394
Rohan Palanikumar, Ahmed A. Elgohary, Simone Martini, Sameh A. Eisa
In this paper, we introduce a model-free, real-time, dynamic optimization and control method for a class of rigid body dynamics. Our method is based on a recent extremum seeking control for vibrational stabilization (ESC-VS) approach that is applicable to a class of second-order mechanical systems. The new ESC-VS method is able to stabilize a rigid body dynamic system about the optimal state of an objective function that can be unknown expression-wise, but assessable through measurements; the ESC-VS is operable by using only one perturbation/vibrational signal. We demonstrate the effectiveness and the applicability of our ESC-VS approach via three rigid-body systems: (1) satellite attitude dynamics, (2) quadcopter attitude dynamics, and (3) acceleration-controlled unicycle dynamics. The results, including simulations with and without measurement delays/noise, illustrate the ability of our ESC-VS to operate successfully as a new methodology of optimization and control for rigid body dynamics.
{"title":"Model-free optimization and control of rigid body dynamics: An extremum seeking for vibrational stabilization approach","authors":"Rohan Palanikumar, Ahmed A. Elgohary, Simone Martini, Sameh A. Eisa","doi":"10.1016/j.ifacsc.2026.100394","DOIUrl":"10.1016/j.ifacsc.2026.100394","url":null,"abstract":"<div><div>In this paper, we introduce a model-free, real-time, dynamic optimization and control method for a class of rigid body dynamics. Our method is based on a recent extremum seeking control for vibrational stabilization (ESC-VS) approach that is applicable to a class of second-order mechanical systems. The new ESC-VS method is able to stabilize a rigid body dynamic system about the optimal state of an objective function that can be unknown expression-wise, but assessable through measurements; the ESC-VS is operable by using <em>only</em> one perturbation/vibrational signal. We demonstrate the effectiveness and the applicability of our ESC-VS approach via three rigid-body systems: (1) satellite attitude dynamics, (2) quadcopter attitude dynamics, and (3) acceleration-controlled unicycle dynamics. The results, including simulations with and without measurement delays/noise, illustrate the ability of our ESC-VS to operate successfully as a new methodology of optimization and control for rigid body dynamics.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"35 ","pages":"Article 100394"},"PeriodicalIF":1.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147384920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-03-01Epub Date: 2026-02-28DOI: 10.1016/j.ifacsc.2026.100399
Johannes Teutsch, Marion Leibold
We present a stochastic data-driven predictive control (DPC) framework for discrete-time linear time-invariant systems subject to sub-Gaussian additive disturbances based solely on input–output data. In contrast to related methods that rely on exact disturbance data or at least sample generation for closed-loop guarantees, the proposed approach leverages a disturbance data estimate. By enforcing consistency of the disturbance data estimate with the available input–output data and system class, we first identify data-driven and provably causal subspace predictors for use in DPC. Then, we analyze statistical properties of the corresponding prediction error, yielding tightened constraints for the nominal predictions that guarantee satisfaction of chance constraints. The proposed DPC scheme comes with guarantees on recursive feasibility and conditional chance constraint satisfaction in closed-loop under standard assumptions. A numerical evaluation study demonstrates the performance of the proposed controller.
{"title":"Stochastic data-driven predictive control of linear systems with sub-Gaussian disturbances using causal predictors","authors":"Johannes Teutsch, Marion Leibold","doi":"10.1016/j.ifacsc.2026.100399","DOIUrl":"10.1016/j.ifacsc.2026.100399","url":null,"abstract":"<div><div>We present a stochastic data-driven predictive control (DPC) framework for discrete-time linear time-invariant systems subject to sub-Gaussian additive disturbances based solely on input–output data. In contrast to related methods that rely on exact disturbance data or at least sample generation for closed-loop guarantees, the proposed approach leverages a disturbance data estimate. By enforcing consistency of the disturbance data estimate with the available input–output data and system class, we first identify data-driven and provably causal subspace predictors for use in DPC. Then, we analyze statistical properties of the corresponding prediction error, yielding tightened constraints for the nominal predictions that guarantee satisfaction of chance constraints. The proposed DPC scheme comes with guarantees on recursive feasibility and conditional chance constraint satisfaction in closed-loop under standard assumptions. A numerical evaluation study demonstrates the performance of the proposed controller.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"35 ","pages":"Article 100399"},"PeriodicalIF":1.8,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147384921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-09-04DOI: 10.1016/j.ifacsc.2025.100337
Dana Zimmermann, Hans-Michael Kaltenbach
Exercise is an important component for glucose management in type 1 diabetes, but remains challenging for automated insulin delivery systems as altered glucose dynamics are difficult to model explicitly. Glucose monitoring data might enable data-driven approaches for learning these dynamics implicitly. We propose combining model predictive control with a reinforcement learning component to adjust basal insulin infusion rates for exercise. We train our model on a variety of exercise scenarios and demonstrate improved glucose control using two different frameworks. We evaluate how generalizable both frameworks are by personalizing a trained model with a small number of additional individual-specific training episodes.
{"title":"Reducing exercise-related hypoglycemia in automated insulin delivery with reinforcement learning","authors":"Dana Zimmermann, Hans-Michael Kaltenbach","doi":"10.1016/j.ifacsc.2025.100337","DOIUrl":"10.1016/j.ifacsc.2025.100337","url":null,"abstract":"<div><div>Exercise is an important component for glucose management in type 1 diabetes, but remains challenging for automated insulin delivery systems as altered glucose dynamics are difficult to model explicitly. Glucose monitoring data might enable data-driven approaches for learning these dynamics implicitly. We propose combining model predictive control with a reinforcement learning component to adjust basal insulin infusion rates for exercise. We train our model on a variety of exercise scenarios and demonstrate improved glucose control using two different frameworks. We evaluate how generalizable both frameworks are by personalizing a trained model with a small number of additional individual-specific training episodes.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100337"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-10-06DOI: 10.1016/j.ifacsc.2025.100345
Ella F.S. Guy , Jennifer L. Knopp , Lui R. Holder-Pearson , J. Geoffrey Chase
Background and Objective:
Feasible methods to assess respiratory compliance and airway resistance without requiring clinical effort or interrupting normal breathing patterns would decrease the high burden of respiratory testing on healthcare systems. This study aims to provide proof of concept of a novel rapid expiratory occlusion (REO) test in a healthy adult population.
Methods:
REO test data was collected for unassisted spontaneous breaths and a PEEP challenge in N=80 healthy adults. Model-identified compliance and resistance values are compared to physiological expectations and literature.
Results:
Median [min, max] compliance was 0.506 [0.199, 1.562] cmH 2O−1L, and resistance was 1.777 [0.811 2.478] cmH 2OL−1s in initial spontaneous breathing, matching expectations. When PEEP was applied compliance decreased (independent of PEEP level) and resistance increased (proportional to set PEEP).
Conclusions:
This study established proof-of-concept efficacy for a model-based REO method identifying compliance and resistance, and informs device development and testing for clinical populations.
{"title":"Identification of passive respiratory mechanics using Rapid Expiratory Occlusions (REOs)","authors":"Ella F.S. Guy , Jennifer L. Knopp , Lui R. Holder-Pearson , J. Geoffrey Chase","doi":"10.1016/j.ifacsc.2025.100345","DOIUrl":"10.1016/j.ifacsc.2025.100345","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Feasible methods to assess respiratory compliance and airway resistance without requiring clinical effort or interrupting normal breathing patterns would decrease the high burden of respiratory testing on healthcare systems. This study aims to provide proof of concept of a novel rapid expiratory occlusion (REO) test in a healthy adult population.</div></div><div><h3>Methods:</h3><div>REO test data was collected for unassisted spontaneous breaths and a PEEP challenge in N=80 healthy adults. Model-identified compliance and resistance values are compared to physiological expectations and literature.</div></div><div><h3>Results:</h3><div>Median [min, max] compliance was 0.506 [0.199, 1.562] cmH <sub>2</sub>O<sup>−1</sup>L, and resistance was 1.777 [0.811 2.478] cmH <sub>2</sub>OL<sup>−1</sup>s in initial spontaneous breathing, matching expectations. When PEEP was applied compliance decreased (independent of PEEP level) and resistance increased (proportional to set PEEP).</div></div><div><h3>Conclusions:</h3><div>This study established proof-of-concept efficacy for a model-based REO method identifying compliance and resistance, and informs device development and testing for clinical populations.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100345"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-09-03DOI: 10.1016/j.ifacsc.2025.100335
Guilherme C. Duran, Edson K. Ueda, André K. Sato, Thiago C. Martins, Marcos S.G. Tsuzuki
Electrode displacement is a common source of error in Electrical Impedance Tomography (EIT), particularly in long-term or dynamic measurements where stable electrode contact is difficult to maintain. This study proposes a comprehensive machine learning framework to detect, classify, and correct electrode displacements prior to image reconstruction. The approach combines tree-based classifiers—such as XGBoost and LightGBM—and Convolutional Neural Networks (CNNs) to identify both the presence and location of displaced electrodes. These models were evaluated across a series of classification tasks with increasing complexity, demonstrating high accuracy even in scenarios involving multiple simultaneous displacements with different magnitudes. For the rectification of distorted voltage measurements, several Denoising Autoencoder (DAE) configurations were investigated. Electrode-specific DAEs trained on all displacement magnitudes achieved an average Mean Squared Error (MSE) reduction of 84.5%, while shift-based DAEs offered computational efficiency for coarse corrections. A hybrid strategy combining fast shift-based and high-accuracy electrode-specific models proved effective and scalable. The pipeline was validated using both synthetic datasets and real EIT measurements, confirming its robustness and generalization capabilities. These results support the use of learning-based correction schemes to improve the reliability and consistency of EIT in practical applications affected by electrode movement.
{"title":"Advancements in Electrical Impedance Tomography: Addressing electrode displacement with artificial neural networks","authors":"Guilherme C. Duran, Edson K. Ueda, André K. Sato, Thiago C. Martins, Marcos S.G. Tsuzuki","doi":"10.1016/j.ifacsc.2025.100335","DOIUrl":"10.1016/j.ifacsc.2025.100335","url":null,"abstract":"<div><div>Electrode displacement is a common source of error in Electrical Impedance Tomography (EIT), particularly in long-term or dynamic measurements where stable electrode contact is difficult to maintain. This study proposes a comprehensive machine learning framework to detect, classify, and correct electrode displacements prior to image reconstruction. The approach combines tree-based classifiers—such as XGBoost and LightGBM—and Convolutional Neural Networks (CNNs) to identify both the presence and location of displaced electrodes. These models were evaluated across a series of classification tasks with increasing complexity, demonstrating high accuracy even in scenarios involving multiple simultaneous displacements with different magnitudes. For the rectification of distorted voltage measurements, several Denoising Autoencoder (DAE) configurations were investigated. Electrode-specific DAEs trained on all displacement magnitudes achieved an average Mean Squared Error (MSE) reduction of 84.5%, while shift-based DAEs offered computational efficiency for coarse corrections. A hybrid strategy combining fast shift-based and high-accuracy electrode-specific models proved effective and scalable. The pipeline was validated using both synthetic datasets and real EIT measurements, confirming its robustness and generalization capabilities. These results support the use of learning-based correction schemes to improve the reliability and consistency of EIT in practical applications affected by electrode movement.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100335"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145061274","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-09-19DOI: 10.1016/j.ifacsc.2025.100340
K.S. Arunab, Aneesh Mathew
This study examined the relationship between Land Surface Temperature (LST) and various controllable, partially controllable, and uncontrollable factors in the cities of Bangalore and Hyderabad. LST showed significant correlations with geographical coordinates in both cities. Despite these directional differences, both cities exhibited consistent correlations with key environmental factors, including Enhanced Vegetation Index (EVI), Normalized Difference Built-up Index (NDBI), Land Cover (LC), Modified Bareness Index (MBI), slope and Modified Normalized Difference Water Index (MNDWI), highlighting the influence of vegetation and built-up areas on urban heat dynamics. The study further compared continuous and grouped LST representations, revealing that grouped LST data exhibited stronger and more consistent correlations with environmental factors, suggesting the presence of non-linear relationships. Factors such as EVI, LC, MBI, MNDWI, Distance to Bare soil (DBS), and Distance to Built-up (DBU) exhibited stronger correlations with grouped LST, highlighting the complexity of LST interactions across different temperature intervals. Grouped LST in Bangalore showed high correlations with LC (0.95), MBI (−0.941), and EVI (−0.938), while in Hyderabad, the strongest associations were with EVI (−0.965), LC (0.929), and DBS (0.918). The study highlights the importance of selecting appropriate LST representations in model development, as stronger correlations with grouped LST suggest non-linearities and potential threshold effects. The study underscores the critical role of vegetation, water bodies, and urban form in shaping LST patterns, offering valuable insights for urban heat mitigation. The study provides valuable insights for policymakers and climate resilience planners, supporting sustainable urban development and enhanced thermal comfort.
{"title":"Exploring the nexus of surface heat and influencing factors in Hyderabad and Bangalore, India","authors":"K.S. Arunab, Aneesh Mathew","doi":"10.1016/j.ifacsc.2025.100340","DOIUrl":"10.1016/j.ifacsc.2025.100340","url":null,"abstract":"<div><div>This study examined the relationship between Land Surface Temperature (LST) and various controllable, partially controllable, and uncontrollable factors in the cities of Bangalore and Hyderabad. LST showed significant correlations with geographical coordinates in both cities. Despite these directional differences, both cities exhibited consistent correlations with key environmental factors, including Enhanced Vegetation Index (EVI), Normalized Difference Built-up Index (NDBI), Land Cover (LC), Modified Bareness Index (MBI), slope and Modified Normalized Difference Water Index (MNDWI), highlighting the influence of vegetation and built-up areas on urban heat dynamics. The study further compared continuous and grouped LST representations, revealing that grouped LST data exhibited stronger and more consistent correlations with environmental factors, suggesting the presence of non-linear relationships. Factors such as EVI, LC, MBI, MNDWI, Distance to Bare soil (DBS), and Distance to Built-up (DBU) exhibited stronger correlations with grouped LST, highlighting the complexity of LST interactions across different temperature intervals. Grouped LST in Bangalore showed high correlations with LC (0.95), MBI (−0.941), and EVI (−0.938), while in Hyderabad, the strongest associations were with EVI (−0.965), LC (0.929), and DBS (0.918). The study highlights the importance of selecting appropriate LST representations in model development, as stronger correlations with grouped LST suggest non-linearities and potential threshold effects. The study underscores the critical role of vegetation, water bodies, and urban form in shaping LST patterns, offering valuable insights for urban heat mitigation. The study provides valuable insights for policymakers and climate resilience planners, supporting sustainable urban development and enhanced thermal comfort.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100340"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study examines a scenario in which individuals, each belonging to a specific type of group (e.g., organizations), are faced with a two-alternative decision-making task. This decision problem is modeled using a novel mixed logit dynamics incorporating conformity biases and committed minority. The model defines two types of conformity biases: internal bias, referred to as inertia, and external bias, referred to as social coordination. Inertia leads group members to adhere to their own status quo, while social coordination drives individuals toward the social majority. We analyze the social model from a control theoretical perspective, proving that social quasi-consensus is stimulated by committed minorities under a bounded rationality condition. In addition to the theoretical results, hypotheses based on the results are validated through numerical experiments.
{"title":"Stability analysis of mixed logit dynamics with internal/external conformity biases and committed minority","authors":"Tatsuya Miyano , Yuji Ito , Daisuke Inoue , Takeshi Hatanaka","doi":"10.1016/j.ifacsc.2025.100343","DOIUrl":"10.1016/j.ifacsc.2025.100343","url":null,"abstract":"<div><div>This study examines a scenario in which individuals, each belonging to a specific type of group (e.g., organizations), are faced with a two-alternative decision-making task. This decision problem is modeled using a novel mixed logit dynamics incorporating conformity biases and committed minority. The model defines two types of conformity biases: internal bias, referred to as inertia, and external bias, referred to as social coordination. Inertia leads group members to adhere to their own status quo, while social coordination drives individuals toward the social majority. We analyze the social model from a control theoretical perspective, proving that social quasi-consensus is stimulated by committed minorities under a bounded rationality condition. In addition to the theoretical results, hypotheses based on the results are validated through numerical experiments.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100343"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate ultra-short-term photovoltaic (PV) power forecasting is essential for grid management and the integration of renewable energy. However, the stochastic and volatile nature of PV power, along with inherent uncertainty, challenges stable grid operation as PV penetration grows. Currently, deep learning (DL) and reinforcement learning (RL) models often struggle to generalize under new conditions, manage computational demands, and address the uncertainty in PV forecasting. To address these issues, a window-based machine learning (WBML) approach is proposed, utilizing light gradient boosting machine (WB-LGBM) and extreme gradient boosting (WB-XGBoost) models. These proposed models outperform attention-based and non-attention-based RL and DL baselines in deterministic metrics like mean absolute error (MAE) and , while significantly reducing training time. Optimized via Optuna and evaluated using fuzzy C-means clustering, their performance is validated by the Diebold–Mariano test. Uncertainty is assessed using non-parametric kernel density estimation (NPKDE) and confidence intervals (CIs) at 99%, 95%, 90%, and 80% confidence levels within the WBML framework, demonstrating robust and conservative forecast uncertainty quantification. Amplitude and phase errors are analyzed with standard deviation error, bias, dispersion, skewness, and kurtosis. The models demonstrate reduced imbalance penalties and enhanced revenue through improved forecasting accuracy.
{"title":"WBML-PV: Window-based machine learning for ultra-short-term photovoltaic power forecasting","authors":"Syed Kumail Hussain Naqvi , Kil To Chong , Hilal Tayara","doi":"10.1016/j.ifacsc.2025.100342","DOIUrl":"10.1016/j.ifacsc.2025.100342","url":null,"abstract":"<div><div>Accurate ultra-short-term photovoltaic (PV) power forecasting is essential for grid management and the integration of renewable energy. However, the stochastic and volatile nature of PV power, along with inherent uncertainty, challenges stable grid operation as PV penetration grows. Currently, deep learning (DL) and reinforcement learning (RL) models often struggle to generalize under new conditions, manage computational demands, and address the uncertainty in PV forecasting. To address these issues, a window-based machine learning (WBML) approach is proposed, utilizing light gradient boosting machine (WB-LGBM) and extreme gradient boosting (WB-XGBoost) models. These proposed models outperform attention-based and non-attention-based RL and DL baselines in deterministic metrics like mean absolute error (MAE) and <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span>, while significantly reducing training time. Optimized via Optuna and evaluated using fuzzy C-means clustering, their performance is validated by the Diebold–Mariano test. Uncertainty is assessed using non-parametric kernel density estimation (NPKDE) and confidence intervals (CIs) at 99%, 95%, 90%, and 80% confidence levels within the WBML framework, demonstrating robust and conservative forecast uncertainty quantification. Amplitude and phase errors are analyzed with standard deviation error, bias, dispersion, skewness, and kurtosis. The models demonstrate reduced imbalance penalties and enhanced revenue through improved forecasting accuracy.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100342"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-09-16DOI: 10.1016/j.ifacsc.2025.100339
Teketel Ketema , Surafel Luleseged Tilahun , Simon D. Zawka , Abebe Geletu
This paper proposes a deep Koopman-based reachability analysis technique for a data-driven control of unknown nonlinear systems subject to process and measurement noises. An intelligent approach combining a neural network and Q-learning algorithm is employed. In particular, the power of the Long Short-Term Memory (LSTM) neural network is leveraged to lift the original nonlinear system into a higher-dimensional space, where the nonlinear dynamics can be approximated linearly, relying solely on the input–output data. The LSTM is set to draw learning insights from Extended Dynamic Mode Decomposition (EDMD) and Information-Theoretic Metric Function (ITMF) results. The Q-learning algorithm is employed to compute adaptive input–output references in the implementation of an adaptive nonlinear zonotopic predictive control technique to compute a robust control input of the system. We also introduced controllability and observability criteria in the presence of noisy data. Finally, a numerical example is given to verify the proposed approach.
{"title":"Deep Koopman-based reachability analysis for data-driven predictive control of unknown nonlinear systems","authors":"Teketel Ketema , Surafel Luleseged Tilahun , Simon D. Zawka , Abebe Geletu","doi":"10.1016/j.ifacsc.2025.100339","DOIUrl":"10.1016/j.ifacsc.2025.100339","url":null,"abstract":"<div><div>This paper proposes a deep Koopman-based reachability analysis technique for a data-driven control of unknown nonlinear systems subject to process and measurement noises. An intelligent approach combining a neural network and Q-learning algorithm is employed. In particular, the power of the Long Short-Term Memory (LSTM) neural network is leveraged to lift the original nonlinear system into a higher-dimensional space, where the nonlinear dynamics can be approximated linearly, relying solely on the input–output data. The LSTM is set to draw learning insights from Extended Dynamic Mode Decomposition (EDMD) and Information-Theoretic Metric Function (ITMF) results. The Q-learning algorithm is employed to compute adaptive input–output references in the implementation of an adaptive nonlinear zonotopic predictive control technique to compute a robust control input of the system. We also introduced controllability and observability criteria in the presence of noisy data. Finally, a numerical example is given to verify the proposed approach.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100339"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145097552","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01Epub Date: 2025-09-09DOI: 10.1016/j.ifacsc.2025.100338
Khadija El Harouri , Soumia El Hani , Nisrine Naseri , Imade Aboudrar , Amina Daghouri
Electric vehicles (EVs) are becoming a basis of sustainable mobility, requiring efficient charging management to minimize costs, balance grid demand, and optimize renewable energy utilization. In workplace parking lots, integrating solar energy and vehicle-to-grid (V2G) technology offers new opportunities for smart energy management. This paper presents an optimization-based charging strategy using Particle Swarm Optimization (PSO) to minimize total energy costs while reducing peak power drawn from the grid, maximizing the use of photovoltaic (PV) energy and ensure that all vehicles reach their target State of Charge (SOC) before leaving the parking lot. Additionally, The proposed approach leverages advantage of V2G technology, enabling EVs to return energy to the grid during peak demand hours, which enhances grid stability and reducing overall energy expenses. A key contribution of this work is the comparative analysis of EV charging management in three different geographical contexts: Morocco, France, and Tunisia. Each country provides distinct energy cost structures, solar availability. A dynamic electricity pricing model is incorporated to adapt the charging strategy based on daily and seasonal tariff variations. The optimization strategy considers multiple constraints like EV arriving and leaving periods, initial and target SOC, PV energy production, and dynamic electricity pricing. Results from simulations indicate that the suggested PSO-based charging strategy achieves significant cost savings can reach up to 65% compared to a conventional unmanaged scenario, reduces peak power coming from the grid, and maximize PV power utilization via self-consumption. Additionally, the findings highlight the benefits of multi-objective optimization in smart parking energy management.
{"title":"Optimizing electric vehicle charging in smart parking lots using particle swarm optimization: A comparative study in Morocco, France, and Tunisia","authors":"Khadija El Harouri , Soumia El Hani , Nisrine Naseri , Imade Aboudrar , Amina Daghouri","doi":"10.1016/j.ifacsc.2025.100338","DOIUrl":"10.1016/j.ifacsc.2025.100338","url":null,"abstract":"<div><div>Electric vehicles (EVs) are becoming a basis of sustainable mobility, requiring efficient charging management to minimize costs, balance grid demand, and optimize renewable energy utilization. In workplace parking lots, integrating solar energy and vehicle-to-grid (V2G) technology offers new opportunities for smart energy management. This paper presents an optimization-based charging strategy using Particle Swarm Optimization (PSO) to minimize total energy costs while reducing peak power drawn from the grid, maximizing the use of photovoltaic (PV) energy and ensure that all vehicles reach their target State of Charge (SOC) before leaving the parking lot. Additionally, The proposed approach leverages advantage of V2G technology, enabling EVs to return energy to the grid during peak demand hours, which enhances grid stability and reducing overall energy expenses. A key contribution of this work is the comparative analysis of EV charging management in three different geographical contexts: Morocco, France, and Tunisia. Each country provides distinct energy cost structures, solar availability. A dynamic electricity pricing model is incorporated to adapt the charging strategy based on daily and seasonal tariff variations. The optimization strategy considers multiple constraints like EV arriving and leaving periods, initial and target SOC, PV energy production, and dynamic electricity pricing. Results from simulations indicate that the suggested PSO-based charging strategy achieves significant cost savings can reach up to 65% compared to a conventional unmanaged scenario, reduces peak power coming from the grid, and maximize PV power utilization via self-consumption. Additionally, the findings highlight the benefits of multi-objective optimization in smart parking energy management.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100338"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145061273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号