Meiling Yue , Xiaohan Si , Jianwen Meng , Yupeng Wang
{"title":"Predicting fuel cell aging behavior under dynamic load through physics-informed machine learning for powertrain system health monitoring","authors":"Meiling Yue , Xiaohan Si , Jianwen Meng , Yupeng Wang","doi":"10.1016/j.conengprac.2025.106617","DOIUrl":"10.1016/j.conengprac.2025.106617","url":null,"abstract":"<div><div>Accurate prediction of fuel cell aging behavior under dynamic load is essential for effective health monitoring and reliable operation in fuel cell powertrain systems. This paper introduces a novel physics-informed machine learning approach to predict future voltage evolution by integrating fuel cell domain knowledge. A physics-based model is integrated into a sequence-to-parameter neural network, predicting parameters used to reconstruct the voltage degradation curve. The prediction process is constrained by data-driven and physics-based loss functions. Experimental results demonstrate that the proposed method achieves a root mean square error of approximately 0.02. Hyperparameter analysis further confirms the robustness of the method. This study offers an effective tool for fuel cell aging prognosis and is expected to be generalized to on-board health monitoring systems within fuel cell powertrain applications.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"166 ","pages":"Article 106617"},"PeriodicalIF":4.6,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145340547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fawang Zhang , Sichao Wu , Yimiao Zhang , Hui Liu , Shida Nie , Jingliang Duan , Rui Liu , Changle Xiang
{"title":"Off-road distributed-drive electric vehicle trajectory tracking control with constrained Transformer MPC","authors":"Fawang Zhang , Sichao Wu , Yimiao Zhang , Hui Liu , Shida Nie , Jingliang Duan , Rui Liu , Changle Xiang","doi":"10.1016/j.conengprac.2025.106608","DOIUrl":"10.1016/j.conengprac.2025.106608","url":null,"abstract":"<div><div>Trajectory tracking control for off-road distributed-drive electric vehicles presents significant challenges due to compound slope effects and rollover risks. Existing approaches often neglect the critical impact of coupled slopes on vehicle roll dynamics and face substantial computational burdens during real-time implementation. This paper presents a four-degree-of-freedom vehicle dynamic model that comprehensively captures longitudinal, lateral, yaw, and roll motions while accounting for compound grade effects. We propose a novel Constrained Transformer Model Predictive Control algorithm that enables real-time policy computation while maintaining safety constraints. CarSim co-simulations demonstrate that our approach effectively prevents rollover and improves trajectory tracking accuracy by 72.28% across various off-road scenarios while reducing computational complexity by 213 times compared to conventional online optimization MPC. Real vehicle tests in off-road environments further validate the effectiveness of the proposed algorithm.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106608"},"PeriodicalIF":4.6,"publicationDate":"2025-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320282","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Boyuan Mu , Chunjun Chen , Kexin Yin , Lu Yang , Yutao Xia
{"title":"Interior pressure control in high-speed trains via curriculum SAC: Simulation and experimental validation","authors":"Boyuan Mu , Chunjun Chen , Kexin Yin , Lu Yang , Yutao Xia","doi":"10.1016/j.conengprac.2025.106622","DOIUrl":"10.1016/j.conengprac.2025.106622","url":null,"abstract":"<div><div>Leveraging the inherent morphological similarities observed in tunnel pressure waveforms, this study proposes a curriculum-based soft actor–critic (CSAC) reinforcement learning algorithm aimed at optimizing the control of air pressure fluctuations inside high-speed trains traversing tunnels. Traditional control strategies frequently struggle to adapt effectively to the dynamic and unpredictable variations inherent in aerodynamic disturbances caused by tunnel environments, resulting in reduced passenger comfort and potential structural fatigue. Addressing this issue, the proposed CSAC algorithm systematically accommodates variations in the amplitude and duration of pressure disturbances by implementing structured curriculum-learning phases, thus progressively guiding the learning process from simple scenarios toward increasingly complex and realistic conditions. Comprehensive comparative analyses were performed using both numerical simulations and scaled-model experiments to validate the effectiveness of the CSAC algorithm. Results consistently demonstrate superior performance, indicating significant reductions in internal pressure fluctuations compared to conventional methods. Furthermore, robustness analyses under uncertain aerodynamic conditions highlight the algorithm’s adaptive capabilities, ensuring reliable performance even when confronted with unforeseen disturbances. Consequently, this curriculum-based reinforcement learning approach provides substantial improvements in passenger comfort and enhances the operational stability and safety of high-speed rail transport systems, presenting a significant advancement in the field of train interior pressure control under complex aerodynamic conditions.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106622"},"PeriodicalIF":4.6,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Self-tuning RLS based multirate adaptive robust position control: A tilting-table test-bench study","authors":"Chenyu Ge , Binh Minh Nguyen , Hiroshi Fujimoto , Masataka Sakamoto , Yuki Terada","doi":"10.1016/j.conengprac.2025.106613","DOIUrl":"10.1016/j.conengprac.2025.106613","url":null,"abstract":"<div><div>Multirate adaptive robust controllers (MARC) have been shown to be a promising position control candidate for high-accuracy machine tools. However, its practical application is still restricted due to the lack of nonlinearity compensation capability as well as the slow convergence speed and limited accuracy of its traditional recursive least-square (RLS) algorithm. To address the above issues, this paper proposes a self-tuning multiple forgetting factor RLS algorithm, which simultaneously identifies the parameters of the linear and nonlinear parts of the system dynamics. Consequently, an adaptive multirate nonlinear compensator was designed to enhance position control performance. The effectiveness of the proposal was evaluated using a qualified test-bench developed in our collaboration with DMG MORI Co., Ltd. Both simulations and experiments confirmed the improved convergence speed and accuracy of the parameter estimation. In particular, the experimental results showed that the root mean square error of position tracking can be reduced by 12. 1% compared to the previous research.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106613"},"PeriodicalIF":4.6,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jinhe Yang , Tongjian Guo , Yi Yu , Peng Jiang , Ce Cao , Xinyu Pang , Quanliang Dong , Xiaoming Wang
{"title":"Adaptive fuzzy fractional-order terminal sliding mode control using adaptive neural network observation architecture with application to ultra-precision stages","authors":"Jinhe Yang , Tongjian Guo , Yi Yu , Peng Jiang , Ce Cao , Xinyu Pang , Quanliang Dong , Xiaoming Wang","doi":"10.1016/j.conengprac.2025.106623","DOIUrl":"10.1016/j.conengprac.2025.106623","url":null,"abstract":"<div><div>This article proposes an Adaptive Fuzzy Fractional-Order Nonsingular Fast Terminal Sliding Mode Control (AFFONFTSMC) strategy for motion control of an ultra-precision magnetic levitation stage (MLS), addressing requirements for exceedingly high trajectory tracking accuracy and strong robustness. The control framework is constructed by integrating a fractional-order sliding surface, an Adaptive Fuzzy Gain (AFG) mechanism, and an online Adaptive Neural Network Observer (ANNO) to achieve precise, real-time estimation of the system’s kinematic states and dynamic parameters, while effectively compensating for model uncertainties and external disturbances. By invoking the Lyapunov stability theorem, adaptive laws are rigorously derived to adjust control parameters online, thereby attenuating chattering phenomena without relying on prior knowledge of uncertainty bounds. Through the ANNO’s continuous dynamic tuning and adaptive parameter compensation, the system rapidly generates optimal control signals, markedly improving both response speed and robustness. Experimental results on the MLS platform demonstrate that the proposed AFFONFTSMC scheme achieves high precision trajectory tracking and exhibits significant advantages in parameter adaptability and disturbance rejection.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106623"},"PeriodicalIF":4.6,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320285","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Lizhou Fang , Kun Zhang , Zhenyu Lu , Jinyuan Liu , Junhui Zhang , Huaizhi Zong , Bing Xu
{"title":"Adaptive robust joint force control for rapid motion of hydraulic limb leg units","authors":"Lizhou Fang , Kun Zhang , Zhenyu Lu , Jinyuan Liu , Junhui Zhang , Huaizhi Zong , Bing Xu","doi":"10.1016/j.conengprac.2025.106596","DOIUrl":"10.1016/j.conengprac.2025.106596","url":null,"abstract":"<div><div>Dynamic locomotion in hydraulic legged robots requires precise and robust joint force control, which is difficult to achieve due to strong system nonlinearities and model uncertainties, especially under high-speed motion conditions. To address these challenges, this paper proposes a Fast Direct Adaptive Robust Control (FDARC) strategy that introduces estimation of uncertain nonlinear nominal parameters, thereby enhancing transient response and tracking performance. Theoretical stability analysis and simulation results verify the effectiveness of the proposed method. Compared with conventional PID and traditional DARC controllers, FDARC reduces joint force tracking error by up to 50.98% (0.5 Hz) and 21.46% (2 Hz) in single-joint experiments. Acting as an inner force control loop, FDARC is further applied to a two-degree-of-freedom hydraulic limb leg unit for vertical jumping experiments. Compared with a feedforward PID (FFPID, as adopted by IIT) and traditional DARC, FDARC enables higher jumping heights (by 6.97% and 4.13%) and significantly reduces foot-end impact forces upon landing (by 27.47% and 43.73%). These results demonstrate the effectiveness of FDARC in achieving fast and robust joints force control, highlighting its potential to enhance the dynamic performance of practical legged robots.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106596"},"PeriodicalIF":4.6,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abu Jafar Md Muzahid , Yang Shi , Zejiang Wang , Anye Zhou , Adian Cook , Chieh Ross Wang , Zhenbo Wang
{"title":"Optimizing on-ramp merging for connected and automated vehicles: A hierarchical approach using deep reinforcement learning and optimal control","authors":"Abu Jafar Md Muzahid , Yang Shi , Zejiang Wang , Anye Zhou , Adian Cook , Chieh Ross Wang , Zhenbo Wang","doi":"10.1016/j.conengprac.2025.106614","DOIUrl":"10.1016/j.conengprac.2025.106614","url":null,"abstract":"<div><div>On-ramp merging for Connected and Automated Vehicles (CAVs) presents significant challenges in dynamic traffic environments. Traditional methods and recent learning-based approaches often fail to simultaneously address decision-making complexity and execution precision under fluctuating conditions. This study introduces a novel hierarchical framework that combines: (1) a high-level Deep Reinforcement Learning (DRL) module that coordinates merging sequences through Virtual Traffic Signals (VTS) with Yield/Green phases and (2) a low-level optimal controller generating collision-free speed trajectories via pseudospectral convex optimization. A convolutional autoencoder compresses high-dimensional traffic states to enhance responsiveness. Extensive simulations demonstrate a 12.5% improvement in mainline throughput a 28% reduction in emergency braking events, and 31.66% lower fuel consumption compared to baseline methods. The framework’s effectiveness in coordinating CAV merges highlights its potential for real-world deployment. Future work will extend validation to multi-lane scenarios with mixed traffic and large-scale multiple merging points.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106614"},"PeriodicalIF":4.6,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hongjun Shao , Donghang Li , Ying Wang , Hongjun Duan , Tao Liu , Baoshan Zhao
{"title":"Current-sensorless modulated model predictive control for LC-filtered four-leg inverters","authors":"Hongjun Shao , Donghang Li , Ying Wang , Hongjun Duan , Tao Liu , Baoshan Zhao","doi":"10.1016/j.conengprac.2025.106619","DOIUrl":"10.1016/j.conengprac.2025.106619","url":null,"abstract":"<div><div>A current-sensorless modulated model predictive control (M2PC) scheme is proposed for LC-filtered four-leg voltage source inverters (VSIs). The generalized parameter estimation-based observer (GPEBO) is an advanced state-observation technique for replacing sensors. First, a current estimator based on GPEBO is designed for LC-filtered four-leg inverters using the M2PC scheme to replace all the current sensors. Hence, the proposed method only requires three voltage sensors, whereas the conventional M2PC scheme requires more sensors. Next, the cost function-based evaluation frame is improved to directly optimize the duty cycle of the candidate control set with pre-selected size, thus the difficulty in solving the exhaustive search and duty cycle optimization of the M2PC scheme for topologies with numerous vectors is overcome. Then, a two-step duty cycle optimization strategy for the M2PC scheme, which can obtain the optimal duty cycles of multiple voltage vectors online, is proposed to improve the tracking accuracy of output voltage. Comparative simulation and experimental results with the advanced M2PC and the improved current-sensorless finite-control-set model predictive control validate the effectiveness of the proposed method.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106619"},"PeriodicalIF":4.6,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Quanxi Zhan , Runjie Shen , Yanmin Zhou , Junrui Zhang , Chenyang Sun , Fenghe Guo , Bin He
{"title":"Multimodal adaptive anti-disturbance control for UAV trajectory tracking in hydropower penstock inspections","authors":"Quanxi Zhan , Runjie Shen , Yanmin Zhou , Junrui Zhang , Chenyang Sun , Fenghe Guo , Bin He","doi":"10.1016/j.conengprac.2025.106615","DOIUrl":"10.1016/j.conengprac.2025.106615","url":null,"abstract":"<div><div>Unmanned aerial vehicles (UAVs) inspecting hydropower penstocks face significant stability challenges due to complex wind fields, including mean winds, shear effects, and turbulence. This paper proposes a Multimodal Adaptive Anti-Disturbance Composite Control (MAADC) framework to enhance trajectory tracking accuracy and disturbance rejection in such environments. The MAADC integrates three core innovations: (1) a Multimodal Adaptive State Compensation Observer (MASCO) for real-time wind disturbance estimation using dual adaptive mechanisms, (2) a Frequency-Domain Wind Disturbance Decomposer (FDWDD) to decouple low- and high-frequency wind components via spectral separation, and (3) a Fast Nonsingular Terminal Sliding Mode Controller (FNTSMC) combining feedforward compensation and adaptive feedback for robust tracking. Experiments in both indoor pipelines and actual hydropower penstocks validate MAADC’s superiority over existing methods. Results show a 58.2% reduction in position RMSE compared to ADRC and 63.9% improvement over AFONFTSMC in turbulent conditions. The framework effectively mitigates wind-induced deviations, achieving sub-0.17 m RMSE in trajectory tracking under dynamic wind shear and turbulence, proving its viability for industrial inspection applications.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106615"},"PeriodicalIF":4.6,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Fixed-time sliding mode control of electro-hydrostatic actuators with adaptive neural network compensation","authors":"Jiahua Ma , Zhikai Yao , Jianyong Yao","doi":"10.1016/j.conengprac.2025.106616","DOIUrl":"10.1016/j.conengprac.2025.106616","url":null,"abstract":"<div><div>This paper presents a new fixed-time sliding mode controller in order to best support high-accuracy motion control of electro-hydrostatic actuators (EHAs). The proposed controller also employs two adaptive neural networks to respectively compensate for unmodeled mechanical dynamics and hydraulic dynamics, thereby reducing the conservatism in control design. A Lyapunov-based stability analysis shows that the controller guarantees practical fixed-time stability in the presence of bounded disturbances. Extensive comparative experiments validate the effectiveness of the approach, demonstrating superior performance and highlighting its potential for high-accuracy EHA motion control applications.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"165 ","pages":"Article 106616"},"PeriodicalIF":4.6,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145320103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}