This study proposes a dual-rate, data-driven system for automated ergometer load adjustment using Heart Rate (HR) and Heart Rate Variability (HRV). The system continuously collects HR and HRV data during exercise to estimate the user’s real-time physiological state and dynamically adjust resistance, maintaining exercise intensity tailored to individual responses. Validation with human participants demonstrated improved HRV without compromising HR tracking performance, highlighting the potential of this approach for personalized training in clinical rehabilitation, athlete conditioning, and general fitness.
{"title":"Motion data-driven exercise design for the simultaneous enhancement of physical capability and psychological state","authors":"Takao Sato, Yoshiharu Kawahara, Natsuki Kawaguchi, Yusuke Tsunoda","doi":"10.1016/j.ifacsc.2025.100349","DOIUrl":"10.1016/j.ifacsc.2025.100349","url":null,"abstract":"<div><div>This study proposes a dual-rate, data-driven system for automated ergometer load adjustment using Heart Rate (HR) and Heart Rate Variability (HRV). The system continuously collects HR and HRV data during exercise to estimate the user’s real-time physiological state and dynamically adjust resistance, maintaining exercise intensity tailored to individual responses. Validation with human participants demonstrated improved HRV without compromising HR tracking performance, highlighting the potential of this approach for personalized training in clinical rehabilitation, athlete conditioning, and general fitness.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"35 ","pages":"Article 100349"},"PeriodicalIF":1.8,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738701","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 : 2025-11-10DOI: 10.1016/j.ifacsc.2025.100348
Xinwei Sun , Lei Zhang
Investigating vibration signals in complex electromechanical systems is essential for improving system stability and control performance. This study proposes a data–physics dual-driven framework to model the dynamic coupling between suspension current and levitation gap in maglev systems. A joint time–frequency analysis is first conducted using Fourier transform, ripple coefficient evaluation, and hysteresis correlation to quantify nonlinear coupling strength and identify a positively lagged relationship between current and gap. To capture this effect, we develop a physics-informed neural network (PINN) that integrates a lag compensation module, embeds electromagnetic equations as physical constraints, and employs an LSTM architecture for end-to-end vibration signal prediction. Unlike conventional approaches that design neural controllers from a control perspective, our method focuses on learning intrinsic coupling patterns directly from real-world operational data. This data-informed modeling approach, enhanced with time-delay compensation and physical consistency, enables accurate prediction of dynamic responses under realistic disturbances. Experiments on data from the Changsha medium-low-speed maglev train show that our model achieves the lowest MAE and RMSE compared to standard PINNs and purely data-driven baselines. It also responds rapidly to gap changes, with a response time of 0.167 ms, making it suitable for real-time maglev control applications. The implementation code is available at: https://github.com/sunning2024/RPinn.
{"title":"A physics-informed LSTM framework with lag compensation for coupled vibration signal modeling","authors":"Xinwei Sun , Lei Zhang","doi":"10.1016/j.ifacsc.2025.100348","DOIUrl":"10.1016/j.ifacsc.2025.100348","url":null,"abstract":"<div><div>Investigating vibration signals in complex electromechanical systems is essential for improving system stability and control performance. This study proposes a data–physics dual-driven framework to model the dynamic coupling between suspension current and levitation gap in maglev systems. A joint time–frequency analysis is first conducted using Fourier transform, ripple coefficient evaluation, and hysteresis correlation to quantify nonlinear coupling strength and identify a positively lagged relationship between current and gap. To capture this effect, we develop a physics-informed neural network (PINN) that integrates a lag compensation module, embeds electromagnetic equations as physical constraints, and employs an LSTM architecture for end-to-end vibration signal prediction. Unlike conventional approaches that design neural controllers from a control perspective, our method focuses on learning intrinsic coupling patterns directly from real-world operational data. This data-informed modeling approach, enhanced with time-delay compensation and physical consistency, enables accurate prediction of dynamic responses under realistic disturbances. Experiments on data from the Changsha medium-low-speed maglev train show that our model achieves the lowest MAE and RMSE compared to standard PINNs and purely data-driven baselines. It also responds rapidly to gap changes, with a response time of 0.167 ms, making it suitable for real-time maglev control applications. The implementation code is available at: <span><span>https://github.com/sunning2024/RPinn</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100348"},"PeriodicalIF":1.8,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145568440","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 : 2025-10-20DOI: 10.1016/j.ifacsc.2025.100341
Javier Urquizo, Patricia Pasmay, Luis Muñoz, Luis Galarza
Permanent magnet synchronous motors play a critical role in modern applications, particularly in the electrification of transportation. Their high energy efficiency and ability to maintain constant power over a wide speed range make them ideal for high-speed trains and electric vehicles. This research explores advanced control strategies, including Field oriented control (FOC), voltage droop control (Vdroop), and dispatchable virtual oscillator control (dVOC), implemented using the Texas Instruments microcontroller development kit, the Boost inverter, and the conventional platform. Furthermore, supervised machine learning algorithms such as support vector machine and reinforcement learning to learn the optimal action-selection policy for an agent interacting with an environment, such as Q-Learning. Large Language Model Meta Artificial Intelligence instruct (LLAMA3) is employed to dynamically optimize control strategies. Laboratory tests validate the implementation, focusing on system efficiency, adaptability, and stability under varying operating conditions. Our findings highlight the potential of artificial intelligence (AI) selected control methods over traditional strategies to deliver optimal performance for modern Permanent magnet synchronous motor.
{"title":"Model-based and Large Language Model Meta Artificial Intelligence techniques for intelligent permanent magnet synchronous motor drive control","authors":"Javier Urquizo, Patricia Pasmay, Luis Muñoz, Luis Galarza","doi":"10.1016/j.ifacsc.2025.100341","DOIUrl":"10.1016/j.ifacsc.2025.100341","url":null,"abstract":"<div><div>Permanent magnet synchronous motors play a critical role in modern applications, particularly in the electrification of transportation. Their high energy efficiency and ability to maintain constant power over a wide speed range make them ideal for high-speed trains and electric vehicles. This research explores advanced control strategies, including Field oriented control (FOC), voltage droop control (Vdroop), and dispatchable virtual oscillator control (dVOC), implemented using the Texas Instruments microcontroller development kit, the Boost inverter, and the conventional platform. Furthermore, supervised machine learning algorithms such as support vector machine and reinforcement learning to learn the optimal action-selection policy for an agent interacting with an environment, such as Q-Learning. Large Language Model Meta Artificial Intelligence instruct (LLAMA3) is employed to dynamically optimize control strategies. Laboratory tests validate the implementation, focusing on system efficiency, adaptability, and stability under varying operating conditions. Our findings highlight the potential of artificial intelligence (AI) selected control methods over traditional strategies to deliver optimal performance for modern Permanent magnet synchronous motor.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100341"},"PeriodicalIF":1.8,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145363461","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 : 2025-10-06DOI: 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":"2025-10-06","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 : 2025-10-06DOI: 10.1016/j.ifacsc.2025.100346
Hassan Ouabi, Rachid Lajouad, Mohammed Kissaoui, Abdelmounime El Magri
This study proposes an advanced multi-mode control strategy for a stand-alone photovoltaic (PV) system equipped with a Li-ion battery. The system is designed to cope with weather fluctuations and varying load demands, which can affect battery lifespan and charging efficiency. The proposed multimode control strategy dynamically switches between three modes: Maximum Power Point Tracking (MPPT) maximizes energy extraction under low PV generation, Constant Current (CC) ensuring fast battery charging, and Constant Voltage (CV) to preserve battery health during saturation. An Artificial Neural Network (ANN) is implemented to adaptively generate the PV reference voltage, enhancing system responsiveness to environmental changes. Furthermore, a state observer is designed to deliver accurate values of all battery states like battery’s state of charge (SoC), ensuring optimized performance, longevity, and safety. The effectiveness of the proposed control strategy and observer is validated through MATLAB/Simulink simulations. Finally, a semi-experimental study based on Processor-in-the-Loop (PIL) testing with the eZdsp TMS320F28335 board confirms the robustness and reliability of the system under real operating conditions.
{"title":"Robust control and state of charge estimation for off-grid solar power systems using ANN-based reference voltage generation","authors":"Hassan Ouabi, Rachid Lajouad, Mohammed Kissaoui, Abdelmounime El Magri","doi":"10.1016/j.ifacsc.2025.100346","DOIUrl":"10.1016/j.ifacsc.2025.100346","url":null,"abstract":"<div><div>This study proposes an advanced multi-mode control strategy for a stand-alone photovoltaic (PV) system equipped with a Li-ion battery. The system is designed to cope with weather fluctuations and varying load demands, which can affect battery lifespan and charging efficiency. The proposed multimode control strategy dynamically switches between three modes: Maximum Power Point Tracking (MPPT) maximizes energy extraction under low PV generation, Constant Current (CC) ensuring fast battery charging, and Constant Voltage (CV) to preserve battery health during saturation. An Artificial Neural Network (ANN) is implemented to adaptively generate the PV reference voltage, enhancing system responsiveness to environmental changes. Furthermore, a state observer is designed to deliver accurate values of all battery states like battery’s state of charge (SoC), ensuring optimized performance, longevity, and safety. The effectiveness of the proposed control strategy and observer is validated through MATLAB/Simulink simulations. Finally, a semi-experimental study based on Processor-in-the-Loop (PIL) testing with the eZdsp TMS320F28335 board confirms the robustness and reliability of the system under real operating conditions.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100346"},"PeriodicalIF":1.8,"publicationDate":"2025-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268210","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":"2025-10-03","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}
Pub Date : 2025-10-03DOI: 10.1016/j.ifacsc.2025.100344
Enes Bajrami, Andrea Kulakov, Eftim Zdravevski, Petre Lameski
Background:
This study addresses national-scale energy optimization using deep reinforcement learning. Unlike prior works that rely on simulated environments or synthetic datasets, this research integrates real-world energy indicators, including electricity generation, greenhouse gas emissions, renewable energy share, fossil fuel dependency, and oil consumption. These indicators, sourced from the World Energy Consumption dataset, capture both developed and developing energy systems, enabling the evaluation of intelligent control policies across diverse contexts.
Methodology:
Two advanced algorithms, Proximal Policy Optimization (PPO) and Soft Actor–Critic (SAC), were implemented and trained using PyTorch across multi-phase evaluation runs (300–3000 episodes). Comparative performance analysis was conducted on key metrics: execution speed, action consistency, and reward optimization. A secondary regional analysis focused on contrasting the Balkan and Nordic countries to evaluate algorithm adaptability between highly developed and developing energy infrastructures.
Significant findings:
SAC demonstrated superior computational throughput and policy stability, making it suitable for real-time and resource-constrained environments. PPO exhibited stronger action magnitudes, enabling more assertive control signals for high-impact interventions. Both agents significantly outperformed a rule-based baseline in responsiveness and adaptability. The proposed framework represents a novel contribution by combining deep reinforcement learning with interpretable, country-level energy indicators. Future work will extend the evaluation to additional continents, including Asia, Africa, and South America, to assess global scalability and applicability.
{"title":"A comparative analysis of PPO and SAC algorithms for energy optimization with country-level energy consumption insights","authors":"Enes Bajrami, Andrea Kulakov, Eftim Zdravevski, Petre Lameski","doi":"10.1016/j.ifacsc.2025.100344","DOIUrl":"10.1016/j.ifacsc.2025.100344","url":null,"abstract":"<div><h3>Background:</h3><div>This study addresses national-scale energy optimization using deep reinforcement learning. Unlike prior works that rely on simulated environments or synthetic datasets, this research integrates real-world energy indicators, including electricity generation, greenhouse gas emissions, renewable energy share, fossil fuel dependency, and oil consumption. These indicators, sourced from the World Energy Consumption dataset, capture both developed and developing energy systems, enabling the evaluation of intelligent control policies across diverse contexts.</div></div><div><h3>Methodology:</h3><div>Two advanced algorithms, Proximal Policy Optimization (PPO) and Soft Actor–Critic (SAC), were implemented and trained using PyTorch across multi-phase evaluation runs (300–3000 episodes). Comparative performance analysis was conducted on key metrics: execution speed, action consistency, and reward optimization. A secondary regional analysis focused on contrasting the Balkan and Nordic countries to evaluate algorithm adaptability between highly developed and developing energy infrastructures.</div></div><div><h3>Significant findings:</h3><div>SAC demonstrated superior computational throughput and policy stability, making it suitable for real-time and resource-constrained environments. PPO exhibited stronger action magnitudes, enabling more assertive control signals for high-impact interventions. Both agents significantly outperformed a rule-based baseline in responsiveness and adaptability. The proposed framework represents a novel contribution by combining deep reinforcement learning with interpretable, country-level energy indicators. Future work will extend the evaluation to additional continents, including Asia, Africa, and South America, to assess global scalability and applicability.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100344"},"PeriodicalIF":1.8,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221234","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 : 2025-09-30DOI: 10.1016/j.ifacsc.2025.100347
Niklas Huhs, Niloofar Kalashtari , Jens Kraitl, Christoph Hornberger, Olaf Simanski
Continuous and non-invasive patient monitoring is essential in healthcare, particularly within Ambient Assisted Living (AAL) environments, to enhance safety and acceptance while preserving privacy. This work investigates two complementary approaches for patient monitoring. In the first approach, a Light Detection and Ranging (LiDAR)-based system was developed to detect and track human subjects in a room using a fine-tuned You Only Look Once, version 5 (YOLOv5) deep learning model. Thanks to LiDAR’s precision and depth sensing capabilities, the system enables live tracking of multiple individuals under varying lighting conditions while safeguarding patient privacy. When the position of the patients in the room is known, the second approach is relevant. A neuromorphic camera, which has a more limited field of view in the room, was employed to measure vital signs such as respiration rate and heart rate by capturing subtle chest movements and micro-vibrations induced by blood circulation. A study involving 26 participants was conducted, with measurements taken at distances ranging from 0.5 metres to 2 metres as well as before and after exercise tasks, consisting of light jogging on a treadmill. Reference data were collected using a Powerlab 15T system equipped with a three-point ECG and a respiration belt. The neuromorphic camera-based measurements demonstrated promising accuracy, validating the feasibility of the approach. Overall, these combined systems offer a contact-free, privacy-preserving solution for continuous patient monitoring, addressing challenges such as limited healthcare staffing, infection control, and the need for vital parameter online tracking in AAL environments.
在医疗保健中，特别是在环境辅助生活（AAL）环境中，持续和非侵入性的患者监测对于提高安全性和接受度，同时保护隐私至关重要。这项工作调查了两种互补的病人监测方法。在第一种方法中，开发了基于光探测和测距（LiDAR）的系统，使用经过微调的You Only Look Once， version 5 （YOLOv5）深度学习模型来检测和跟踪房间中的人类受试者。由于激光雷达的精度和深度传感能力，该系统可以在不同的照明条件下实时跟踪多个个体，同时保护患者的隐私。当病人在房间里的位置是已知的，第二种方法是相关的。神经形态相机在房间内的视野更有限，通过捕捉微妙的胸部运动和血液循环引起的微振动来测量呼吸率和心率等生命体征。研究人员对26名参与者进行了研究，测量了他们在0.5米到2米之间的距离，以及在锻炼任务（包括在跑步机上慢跑）之前和之后的运动量。参考数据的收集使用配备有三点心电图和呼吸带的Powerlab 15T系统。基于神经形态相机的测量显示出良好的准确性，验证了该方法的可行性。总的来说，这些组合系统为患者的持续监测提供了一种无接触、保护隐私的解决方案，解决了诸如有限的医疗人员、感染控制以及AAL环境中对重要参数在线跟踪的需求等挑战。
{"title":"Application of LiDAR and neuromorphic vision in Ambient Assisted Living environments","authors":"Niklas Huhs, Niloofar Kalashtari , Jens Kraitl, Christoph Hornberger, Olaf Simanski","doi":"10.1016/j.ifacsc.2025.100347","DOIUrl":"10.1016/j.ifacsc.2025.100347","url":null,"abstract":"<div><div>Continuous and non-invasive patient monitoring is essential in healthcare, particularly within Ambient Assisted Living (AAL) environments, to enhance safety and acceptance while preserving privacy. This work investigates two complementary approaches for patient monitoring. In the first approach, a Light Detection and Ranging (LiDAR)-based system was developed to detect and track human subjects in a room using a fine-tuned You Only Look Once, version 5 (YOLOv5) deep learning model. Thanks to LiDAR’s precision and depth sensing capabilities, the system enables live tracking of multiple individuals under varying lighting conditions while safeguarding patient privacy. When the position of the patients in the room is known, the second approach is relevant. A neuromorphic camera, which has a more limited field of view in the room, was employed to measure vital signs such as respiration rate and heart rate by capturing subtle chest movements and micro-vibrations induced by blood circulation. A study involving 26 participants was conducted, with measurements taken at distances ranging from 0.5 metres to 2 metres as well as before and after exercise tasks, consisting of light jogging on a treadmill. Reference data were collected using a Powerlab 15T system equipped with a three-point ECG and a respiration belt. The neuromorphic camera-based measurements demonstrated promising accuracy, validating the feasibility of the approach. Overall, these combined systems offer a contact-free, privacy-preserving solution for continuous patient monitoring, addressing challenges such as limited healthcare staffing, infection control, and the need for vital parameter online tracking in AAL environments.</div></div>","PeriodicalId":29926,"journal":{"name":"IFAC Journal of Systems and Control","volume":"34 ","pages":"Article 100347"},"PeriodicalIF":1.8,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268211","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":"2025-09-26","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 : 2025-09-19DOI: 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":"2025-09-19","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}
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