Pub Date : 2025-12-31DOI: 10.1016/j.conengprac.2025.106732
Mengfei Zhang , Zhao Feng , Xiaohui Zhang , Xingyu Chen , Deng Li , Xiaohui Xiao
Navigating non-coaxial two-wheeled robots through narrow environments presents dual challenges: ensuring collision-free motion and maintaining lateral safety during aggressive steering maneuvers required for obstacle avoidance. This paper proposes a novel approach which integrates two complementary safety mechanisms within a model predictive control (MPC) architecture. The geometry-enhanced navigation safety control barrier function (GNSCBF) accurately captures the robot’s varying spatial occupancy during steering through a novel articulated geometry representation. The multi-step roll safety control barrier function (MRSCBF) predicts future roll dynamics over a finite horizon and optimizes velocity profiles through quadratic programming (QP) to ensure lateral safety. Experimental validation on real-world platform demonstrates the effectiveness of the proposed approach.
{"title":"Composite safety control for non-coaxial two-wheeled robot with geometry-enhanced multi-step control barrier function","authors":"Mengfei Zhang , Zhao Feng , Xiaohui Zhang , Xingyu Chen , Deng Li , Xiaohui Xiao","doi":"10.1016/j.conengprac.2025.106732","DOIUrl":"10.1016/j.conengprac.2025.106732","url":null,"abstract":"<div><div>Navigating non-coaxial two-wheeled robots through narrow environments presents dual challenges: ensuring collision-free motion and maintaining lateral safety during aggressive steering maneuvers required for obstacle avoidance. This paper proposes a novel approach which integrates two complementary safety mechanisms within a model predictive control (MPC) architecture. The geometry-enhanced navigation safety control barrier function (GNSCBF) accurately captures the robot’s varying spatial occupancy during steering through a novel articulated geometry representation. The multi-step roll safety control barrier function (MRSCBF) predicts future roll dynamics over a finite horizon and optimizes velocity profiles through quadratic programming (QP) to ensure lateral safety. Experimental validation on real-world platform demonstrates the effectiveness of the proposed approach.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106732"},"PeriodicalIF":4.6,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884371","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}
Pub Date : 2025-12-30DOI: 10.1016/j.conengprac.2025.106737
Xiaocong Li , Alejandro Gonzalez-Garcia , Herman Castañeda
Unmanned surface vehicles (USVs) have gained significant attention recently for applications such as delivery and trash removal. However, accurately modeling these vehicles is difficult due to their inherent underactuation and complex dynamics, which often result in inaccurate tracking. To address this challenge, we propose a data-enabled learning approach to fully exploit the abundant data available for achieving enhanced control performance. The core concept is that suboptimal motion generates a substantial amount of data, specifically related to surge, yaw rate, and control inputs. This rich information can enable an efficient learning process to enhance motion control. In this work, we use data collected from experiments to optimize planar motion control in an underactuated vessel. The optimization algorithm allows for efficient tuning of the control gains for a predefined controller, with quick convergence. Importantly, the gain optimization does not require knowledge of the vehicle model. Simulations and experiments conducted on a vessel prototype demonstrate improved controller performance and efficiency in learning.
{"title":"Optimizing unmanned surface vehicle control: A data-enabled learning approach","authors":"Xiaocong Li , Alejandro Gonzalez-Garcia , Herman Castañeda","doi":"10.1016/j.conengprac.2025.106737","DOIUrl":"10.1016/j.conengprac.2025.106737","url":null,"abstract":"<div><div>Unmanned surface vehicles (USVs) have gained significant attention recently for applications such as delivery and trash removal. However, accurately modeling these vehicles is difficult due to their inherent underactuation and complex dynamics, which often result in inaccurate tracking. To address this challenge, we propose a data-enabled learning approach to fully exploit the abundant data available for achieving enhanced control performance. The core concept is that suboptimal motion generates a substantial amount of data, specifically related to surge, yaw rate, and control inputs. This rich information can enable an efficient learning process to enhance motion control. In this work, we use data collected from experiments to optimize planar motion control in an underactuated vessel. The optimization algorithm allows for efficient tuning of the control gains for a predefined controller, with quick convergence. Importantly, the gain optimization does not require knowledge of the vehicle model. Simulations and experiments conducted on a vessel prototype demonstrate improved controller performance and efficiency in learning.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106737"},"PeriodicalIF":4.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884382","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}
Pub Date : 2025-12-30DOI: 10.1016/j.conengprac.2025.106705
Daniel Igbokwe , Malek Ghanes , Marc Bodson , Mohamed Hamida , Amir Messali
This paper introduces a novel discrete sliding-mode control (DSM) strategy for multiphase synchronous machines, supported by an adaptive discrete hybrid filtering differentiator (DHFD). The proposed reaching law significantly enhances transient performance while reducing computational complexity, enabling real-time implementation on low-cost processors. Focusing on five-phase wound-rotor synchronous machines (WRSMs)—a rarely studied configuration in existing literature—the method demonstrates remarkable efficacy in handling transient dynamics and parameter uncertainties. Notably, the control scheme is directly applicable to permanent magnet synchronous machines (PMSMs) and extensible to *n*-phase systems beyond five phases. Experimental hardware-in-the-loop (HIL) and simulation results validate the performance of the approach, showcasing rapid convergence, chattering suppression, and robustness under dynamic loads. By bridging the gap between advanced discrete-time sliding-mode theory and practical implementation for multiphase machines, this work offers a versatile solution for high-performance motor drives in aerospace, automotive, and industrial applications.
{"title":"A novel discrete sliding mode control based adaptive differentiator for five phase synchronous machines","authors":"Daniel Igbokwe , Malek Ghanes , Marc Bodson , Mohamed Hamida , Amir Messali","doi":"10.1016/j.conengprac.2025.106705","DOIUrl":"10.1016/j.conengprac.2025.106705","url":null,"abstract":"<div><div>This paper introduces a novel discrete sliding-mode control (DSM) strategy for multiphase synchronous machines, supported by an adaptive discrete hybrid filtering differentiator (DHFD). The proposed reaching law significantly enhances transient performance while reducing computational complexity, enabling real-time implementation on low-cost processors. Focusing on five-phase wound-rotor synchronous machines (WRSMs)—a rarely studied configuration in existing literature—the method demonstrates remarkable efficacy in handling transient dynamics and parameter uncertainties. Notably, the control scheme is directly applicable to permanent magnet synchronous machines (PMSMs) and extensible to *n*-phase systems beyond five phases. Experimental hardware-in-the-loop (HIL) and simulation results validate the performance of the approach, showcasing rapid convergence, chattering suppression, and robustness under dynamic loads. By bridging the gap between advanced discrete-time sliding-mode theory and practical implementation for multiphase machines, this work offers a versatile solution for high-performance motor drives in aerospace, automotive, and industrial applications.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106705"},"PeriodicalIF":4.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884376","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}
Pub Date : 2025-12-30DOI: 10.1016/j.conengprac.2025.106727
Xu Zhou, Biao Yang, Fengshan Dou, Zhiqiang Long
Maintaining stability under strong aerodynamic disturbances presents a fundamental challenge for magnetic suspension and balance systems (MSBS). This paper presents a data-driven framework for quantifying the closed-loop stability margin of the axial suspension system in MSBS using full-state information, with a focus on stability margin as the key metric for evaluating system robustness against uncertainties. First, a dynamic model of the axial suspension system is established, and the stability margin is analyzed based on coprime factorization theory.Subsequently, an iterative learning-based tracking differentiator (IL-TD) is developed to extract velocity signals and augment system outputs, while a closed-loop reconfiguration strategy maintains the original system’s stability. Thereafter, stable kernel representation (SKR) and stable image representation (SIR) are introduced to enable data-driven implementation of the stability margin. The stability margin is characterized through data-driven identification of SKR and SIR from input-output data. Based on the estimated stability margin, stability performance monitoring is conducted, enabling the classification of closed-loop stability into distinct levels. This model-agnostic approach requires no prior knowledge of system models and has been experimentally validated on a low-speed wind tunnel MSBS platform.
{"title":"A data-driven framework for stability margin estimation in magnetic suspension and balance systems","authors":"Xu Zhou, Biao Yang, Fengshan Dou, Zhiqiang Long","doi":"10.1016/j.conengprac.2025.106727","DOIUrl":"10.1016/j.conengprac.2025.106727","url":null,"abstract":"<div><div>Maintaining stability under strong aerodynamic disturbances presents a fundamental challenge for magnetic suspension and balance systems (MSBS). This paper presents a data-driven framework for quantifying the closed-loop stability margin of the axial suspension system in MSBS using full-state information, with a focus on stability margin as the key metric for evaluating system robustness against uncertainties. First, a dynamic model of the axial suspension system is established, and the stability margin is analyzed based on coprime factorization theory.Subsequently, an iterative learning-based tracking differentiator (IL-TD) is developed to extract velocity signals and augment system outputs, while a closed-loop reconfiguration strategy maintains the original system’s stability. Thereafter, stable kernel representation (SKR) and stable image representation (SIR) are introduced to enable data-driven implementation of the stability margin. The stability margin is characterized through data-driven identification of SKR and SIR from input-output data. Based on the estimated stability margin, stability performance monitoring is conducted, enabling the classification of closed-loop stability into distinct levels. This model-agnostic approach requires no prior knowledge of system models and has been experimentally validated on a low-speed wind tunnel MSBS platform.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106727"},"PeriodicalIF":4.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884287","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}
Pub Date : 2025-12-30DOI: 10.1016/j.conengprac.2025.106730
Wenpeng Wei , Teng Zuo , Huaicheng Yao , Tianyi He
The Electro-Mechanical Brake (EMB) system for electric vehicle exhibits characteristics such as nonlinearity, multistage switching, and parameter variations due to wear and thermal effects, which impose great challenges for accurate modeling and parameters estimation. The novelty of this paper lies in the development of a unified Koopman operator model for clamping force, a hybrid physics-Koopman model that provides linear formulation for nonlinear EMB dynamics, and adaptive parameters estimations for the hybrid model. Firstly, a unified data-driven clamping force Koopman operator model without switching logic that captures the nonlinearity and hysteresis is developed. The Koopman model is then integrated with known physical principles to render an augmented linear representation for the nonlinear EMB system. After that, based on the hybrid Koopman-physics linear model, the adaptive recursive least-square estimation algorithm is implemented to estimate the time-varying system parameters in the real-time. The proposed approach is validated in both simulations and experiments. In the simulation validation, the estimation errors of all system parameters are less than 1.2%. Besides, the motor speed is used to validate the model accuracy, the RMSE and RRMSE of predicted motor speed are found to be 0.9 rad/s and 0.02%, respectively. In the experimental validation, the motor speed is accurately predicted by the hybrid Koopman-physics model with RMSE and RRMSE of 0.52 rad/s and 0.11%. Furthermore, the proposed framework is compared with traditional approaches. The brake force estimation have 41.67% improvement in RMSE and 41.10% improvement in RRMSE compared to the traditional polynomial-based approach. The unified model achieves 58.3% improvement in RMSE and 55.8% improvement in RRMSE in terms of motor speed response compared to the traditional physical model.
{"title":"Unified hybrid-Koopman modeling and adaptive parameters estimation of nonlinear Electro-Mechanical Brake systems","authors":"Wenpeng Wei , Teng Zuo , Huaicheng Yao , Tianyi He","doi":"10.1016/j.conengprac.2025.106730","DOIUrl":"10.1016/j.conengprac.2025.106730","url":null,"abstract":"<div><div>The Electro-Mechanical Brake (EMB) system for electric vehicle exhibits characteristics such as nonlinearity, multistage switching, and parameter variations due to wear and thermal effects, which impose great challenges for accurate modeling and parameters estimation. The novelty of this paper lies in the development of a unified Koopman operator model for clamping force, a hybrid physics-Koopman model that provides linear formulation for nonlinear EMB dynamics, and adaptive parameters estimations for the hybrid model. Firstly, a unified data-driven clamping force Koopman operator model without switching logic that captures the nonlinearity and hysteresis is developed. The Koopman model is then integrated with known physical principles to render an augmented linear representation for the nonlinear EMB system. After that, based on the hybrid Koopman-physics linear model, the adaptive recursive least-square estimation algorithm is implemented to estimate the time-varying system parameters in the real-time. The proposed approach is validated in both simulations and experiments. In the simulation validation, the estimation errors of all system parameters are less than 1.2%. Besides, the motor speed is used to validate the model accuracy, the RMSE and RRMSE of predicted motor speed are found to be 0.9 rad/s and 0.02%, respectively. In the experimental validation, the motor speed is accurately predicted by the hybrid Koopman-physics model with RMSE and RRMSE of 0.52 rad/s and 0.11%. Furthermore, the proposed framework is compared with traditional approaches. The brake force estimation have 41.67% improvement in RMSE and 41.10% improvement in RRMSE compared to the traditional polynomial-based approach. The unified model achieves 58.3% improvement in RMSE and 55.8% improvement in RRMSE in terms of motor speed response compared to the traditional physical model.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106730"},"PeriodicalIF":4.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884377","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}
Pub Date : 2025-12-30DOI: 10.1016/j.conengprac.2025.106734
Jian-Ru Huo , Weiyao Lan , Xiao Yu
This paper investigates the design problem of false data injection (FDI) attacks aimed at degrading system performance in multi-agent systems (MASs). We consider a scenario where a limited number of agents can be compromised, which implies that the attack signal exhibits a sparse structure. However, ensuring this sparsity is challenging because of the high dimensionality of the attack signal. To overcome this difficulty, the attack signal is decomposed into two components: the bias signal and the attack direction vector, leading to a two-stage attack strategy design method. In the first stage, we ensure sparsity by exploiting the lower-dimensional structure of the bias signal. To this end, we formulate an optimization problem with an l0 norm constraint, which is solved using the alternating direction method of multipliers (ADMM). In the second stage, to achieve better attack performance, the attack direction vector is constructed using and H∞ indices in the frequency domain, ensuring both the effectiveness and stealthiness of the attack. Moreover, we derive the ultimate bounds for consensus error and state estimation error induced by the attack. Finally, we provide a hardware-in-the-loop (HIL) simulation in Gazebo platform and a real-world experiment on multiple mobile robots to illustrate the effectiveness of the proposed attack strategy.
{"title":"Sparse false data injection attacks against distributed control of multi-agent systems","authors":"Jian-Ru Huo , Weiyao Lan , Xiao Yu","doi":"10.1016/j.conengprac.2025.106734","DOIUrl":"10.1016/j.conengprac.2025.106734","url":null,"abstract":"<div><div>This paper investigates the design problem of false data injection (FDI) attacks aimed at degrading system performance in multi-agent systems (MASs). We consider a scenario where a limited number of agents can be compromised, which implies that the attack signal exhibits a sparse structure. However, ensuring this sparsity is challenging because of the high dimensionality of the attack signal. To overcome this difficulty, the attack signal is decomposed into two components: the bias signal and the attack direction vector, leading to a two-stage attack strategy design method. In the first stage, we ensure sparsity by exploiting the lower-dimensional structure of the bias signal. To this end, we formulate an optimization problem with an <em>l</em><sub>0</sub> norm constraint, which is solved using the alternating direction method of multipliers (ADMM). In the second stage, to achieve better attack performance, the attack direction vector is constructed using <span><math><msub><mi>H</mi><mo>−</mo></msub></math></span> and <em>H</em><sub>∞</sub> indices in the frequency domain, ensuring both the effectiveness and stealthiness of the attack. Moreover, we derive the ultimate bounds for consensus error and state estimation error induced by the attack. Finally, we provide a hardware-in-the-loop (HIL) simulation in Gazebo platform and a real-world experiment on multiple mobile robots to illustrate the effectiveness of the proposed attack strategy.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106734"},"PeriodicalIF":4.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884373","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}
Pub Date : 2025-12-29DOI: 10.1016/j.conengprac.2025.106726
Yu’ao Wang , Tong Yang , Zhixin Yang , Ming Li , Yongchun Fang , Ning Sun
Pneumatic artificial muscle (PAM)-actuated robots always exhibit notable compliance and satisfactory human-robot interaction performance, while also suffering from complex nonlinear behaviors such as hysteresis and creep, and are highly susceptible to external disturbances. To this end, in this paper, an adaptive controller based on the fully actuated system (FAS) approach is proposed for antagonistic PAM-actuated robotic arms to achieve precise motion control without accurate system models, while maintaining strong robustness against environmental uncertainties. First, an echo state network (ESN) is elaborately integrated to estimate the systems’ unknown dynamics in real time. Then, a super-twisting extended state observer (STESO) utilizing ESN-generated information is designed to further compensate for external disturbances. Under the high-order sliding mode control framework based on FAS approaches, the plant nonlinearities are compensated by the STESO-ESN hybrid algorithm, thereby transforming the system into a closed-loop linear form. As a result, the closed-loop dynamics are flexibly shaped through pole placement, to realize high-precision trajectory tracking. Finally, a rigorous Lyapunov-based stability analysis is provided, and a series of experiments are conducted on a self-built experimental platform to validate the effectiveness of the proposed method.
{"title":"A hybrid echo state network-based fully actuated system approach for antagonistic PAM-actuated robotic arms under external disturbances","authors":"Yu’ao Wang , Tong Yang , Zhixin Yang , Ming Li , Yongchun Fang , Ning Sun","doi":"10.1016/j.conengprac.2025.106726","DOIUrl":"10.1016/j.conengprac.2025.106726","url":null,"abstract":"<div><div>Pneumatic artificial muscle (PAM)-actuated robots always exhibit notable compliance and satisfactory human-robot interaction performance, while also suffering from complex nonlinear behaviors such as hysteresis and creep, and are highly susceptible to external disturbances. To this end, in this paper, an adaptive controller based on the fully actuated system (FAS) approach is proposed for antagonistic PAM-actuated robotic arms to achieve precise motion control without accurate system models, while maintaining strong robustness against environmental uncertainties. First, an echo state network (ESN) is elaborately integrated to estimate the systems’ unknown dynamics in real time. Then, a super-twisting extended state observer (STESO) utilizing ESN-generated information is designed to further compensate for external disturbances. Under the high-order sliding mode control framework based on FAS approaches, the plant nonlinearities are compensated by the STESO-ESN hybrid algorithm, thereby transforming the system into a closed-loop linear form. As a result, the closed-loop dynamics are flexibly shaped through pole placement, to realize high-precision trajectory tracking. Finally, a rigorous Lyapunov-based stability analysis is provided, and a series of experiments are conducted on a self-built experimental platform to validate the effectiveness of the proposed method.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106726"},"PeriodicalIF":4.6,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884375","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}
This paper presents an integrated system for autonomous lane following with obstacle avoidance in unmanned tracked vehicles (UTVs), combining monocular vision and Active Disturbance Rejection Control (ADRC). A vision-based guidance system is developed using deep learning models: YOLOPv2 for lane segmentation and YOLOv8 for obstacle detection within a dynamic region of interest. Novel lane processing algorithms address partial detections and generate aligned lane boundaries, while a computationally efficient virtual lane generation mechanism enables path planning around obstacles without requiring dedicated depth sensors. To follow the path defined by this guidance system, an ADRC controller is designed for the UTV’s lateral control channel based on a kinematic model, incorporating disturbance estimation via an extended state observer, ensuring robust regulation of lateral path error. The system’s effectiveness is demonstrated through comprehensive experimental validation on a physical UTV platform in two distinct environments: an indoor track with static obstacles and an outdoor setting with both static and dynamic obstacles. Outdoor trials confirm the system’s robustness against real-world challenges, including sloped terrain, varying natural lighting, and multi-colored lane markings. Furthermore, the system successfully navigated around obstacles and critically validated its fail-safe stop logic when the path was fully blocked. Comparative tests against a conventional PID controller quantitatively demonstrate the ADRC’s superior tracking accuracy and disturbance rejection capabilities, highlighting its enhanced robustness in both controlled indoor and unstructured outdoor environments. These results confirm the feasibility of achieving robust lane following and effective obstacle avoidance in UTVs using cost-efficient monocular vision. Supplementary material: https://youtu.be/9aKGugeYmfw?si=qiBCTzi7hYUvwUW6
{"title":"Lane following with obstacle avoidance for unmanned tracked vehicles using monocular vision and active disturbance rejection control","authors":"Salem-Bilal Amokrane , Momir Stanković , Rafal Madonski , Benyahia Ahmed Taki-Eddine","doi":"10.1016/j.conengprac.2025.106723","DOIUrl":"10.1016/j.conengprac.2025.106723","url":null,"abstract":"<div><div>This paper presents an integrated system for autonomous lane following with obstacle avoidance in unmanned tracked vehicles (UTVs), combining monocular vision and Active Disturbance Rejection Control (ADRC). A vision-based guidance system is developed using deep learning models: YOLOPv2 for lane segmentation and YOLOv8 for obstacle detection within a dynamic region of interest. Novel lane processing algorithms address partial detections and generate aligned lane boundaries, while a computationally efficient virtual lane generation mechanism enables path planning around obstacles without requiring dedicated depth sensors. To follow the path defined by this guidance system, an ADRC controller is designed for the UTV’s lateral control channel based on a kinematic model, incorporating disturbance estimation via an extended state observer, ensuring robust regulation of lateral path error. The system’s effectiveness is demonstrated through comprehensive experimental validation on a physical UTV platform in two distinct environments: an indoor track with static obstacles and an outdoor setting with both static and dynamic obstacles. Outdoor trials confirm the system’s robustness against real-world challenges, including sloped terrain, varying natural lighting, and multi-colored lane markings. Furthermore, the system successfully navigated around obstacles and critically validated its fail-safe stop logic when the path was fully blocked. Comparative tests against a conventional PID controller quantitatively demonstrate the ADRC’s superior tracking accuracy and disturbance rejection capabilities, highlighting its enhanced robustness in both controlled indoor and unstructured outdoor environments. These results confirm the feasibility of achieving robust lane following and effective obstacle avoidance in UTVs using cost-efficient monocular vision. <em>Supplementary material</em>: <span><span>https://youtu.be/9aKGugeYmfw?si=qiBCTzi7hYUvwUW6</span><svg><path></path></svg></span></div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106723"},"PeriodicalIF":4.6,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884286","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}
Pub Date : 2025-12-27DOI: 10.1016/j.conengprac.2025.106720
Jingtao Hu, Tianrui Chen, Cheng Zhou, Weiming Wu, Cong Wang
This paper presents a learning approach for nonlinear dynamical systems using noisy sampled-output data. A sampled-data observer with a nonlinear gain structure is first designed to reconstruct the state trajectory, achieving both rapid convergence and reduced sensitivity to measurement noise. A deterministic learning process based on the reconstructed trajectory is then employed to identify the system dynamics. By exploiting the partial persistent excitation (PE) property of the radial basis function (RBF) neural network, a generalized exponential convergence model is derived for the perturbed linear time-varying (LTV) system associated with the identification process. This model explicitly relates the observer gain, network structure, and noise level to learning performance, providing theoretical guidance for parameter selection. Furthermore, the learned dynamics are reused to construct a non-high-gain observer, enabling accurate state estimation with low computational complexity in similar tasks. The proposed approach is validated through simulation and compressor aerodynamic instability warning experiments, demonstrating its capability for accurate learning and high-performance utilization of nonlinear dynamics under noisy conditions.
{"title":"Sampled-data observer-based deterministic learning in noisy environments and its performance analysis","authors":"Jingtao Hu, Tianrui Chen, Cheng Zhou, Weiming Wu, Cong Wang","doi":"10.1016/j.conengprac.2025.106720","DOIUrl":"10.1016/j.conengprac.2025.106720","url":null,"abstract":"<div><div>This paper presents a learning approach for nonlinear dynamical systems using noisy sampled-output data. A sampled-data observer with a nonlinear gain structure is first designed to reconstruct the state trajectory, achieving both rapid convergence and reduced sensitivity to measurement noise. A deterministic learning process based on the reconstructed trajectory is then employed to identify the system dynamics. By exploiting the partial persistent excitation (PE) property of the radial basis function (RBF) neural network, a generalized exponential convergence model is derived for the perturbed linear time-varying (LTV) system associated with the identification process. This model explicitly relates the observer gain, network structure, and noise level to learning performance, providing theoretical guidance for parameter selection. Furthermore, the learned dynamics are reused to construct a non-high-gain observer, enabling accurate state estimation with low computational complexity in similar tasks. The proposed approach is validated through simulation and compressor aerodynamic instability warning experiments, demonstrating its capability for accurate learning and high-performance utilization of nonlinear dynamics under noisy conditions.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106720"},"PeriodicalIF":4.6,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884285","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}
Pub Date : 2025-12-27DOI: 10.1016/j.conengprac.2025.106735
Samuel Sonnino , Stefano Melzi , Francesco Pirchio , Pietro Caresia , Alessandro Manzoni , Gianluca Vaini
This study investigates the potential of Active Suspension systems to modulate Roll Stiffness Distribution (RSD) with the goal of enhancing vehicle lateral performance and stability. By shifting the suspension’s role beyond conventional vertical dynamics and ride comfort, widely explored in existing literature, the proposed approach leverages Roll Stiffness Distribution between front and rear axles to actively influence handling balance across a broad range of driving conditions. The approach utilizes actuators already integrated into the suspension systems of modern vehicles, maximizing their potential through a dedicated control strategy. A control logic for the active management of RSD is introduced, designed to mitigate understeer or oversteer behavior in real time. System validation is initially performed in a simulation environment using standardized open-loop maneuvers in accordance with ISO protocols (Ramp Steer, Step Steer, Sine with Dwell). The controller is optimized using a Genetic Algorithm (GA) and subsequently validated through Driver-in-the-Loop (DiL) testing, conducted with a sample of 24 drivers at the dynamic simulator DriSMi of Politecnico di Milano. These tests aimed to assess both objective performance and subjective driver perception. Finally, the proposed system was integrated into a vehicle equipped with an Active Rear Steering system to explore the synergies between the two control systems and define their respective domains of effectiveness in lateral dynamics optimization.
本研究探讨了主动悬架系统调节侧倾刚度分布（RSD）的潜力，目的是提高车辆的横向性能和稳定性。通过将悬架的作用转移到传统的垂直动力学和乘坐舒适性之外，该方法在现有文献中得到了广泛的探讨，该方法利用前后轴之间的侧倾刚度分布，在广泛的驾驶条件下积极影响操纵平衡。该方法利用已经集成到现代车辆悬架系统中的执行器，通过专用控制策略最大限度地发挥其潜力。引入了一种用于主动管理RSD的控制逻辑，旨在实时缓解转向不足或转向过度行为。系统验证最初在模拟环境中进行，使用符合ISO协议的标准化开环机动（斜坡转向，步进转向，正弦与Dwell）。控制器采用遗传算法（GA）进行优化，随后在米兰理工大学（Politecnico di Milano）的动态模拟器DriSMi上对24名驾驶员进行了驾驶员在环（DiL）测试。这些测试旨在评估客观性能和主观驾驶员感知。最后，将所提出的系统集成到配备主动后转向系统的车辆中，以探索两种控制系统之间的协同作用，并确定各自在横向动力学优化中的有效性领域。
{"title":"Active control of vehicle lateral dynamics through roll stiffness distribution: Simulation and driver-in-the-loop testing","authors":"Samuel Sonnino , Stefano Melzi , Francesco Pirchio , Pietro Caresia , Alessandro Manzoni , Gianluca Vaini","doi":"10.1016/j.conengprac.2025.106735","DOIUrl":"10.1016/j.conengprac.2025.106735","url":null,"abstract":"<div><div>This study investigates the potential of Active Suspension systems to modulate Roll Stiffness Distribution (RSD) with the goal of enhancing vehicle lateral performance and stability. By shifting the suspension’s role beyond conventional vertical dynamics and ride comfort, widely explored in existing literature, the proposed approach leverages Roll Stiffness Distribution between front and rear axles to actively influence handling balance across a broad range of driving conditions. The approach utilizes actuators already integrated into the suspension systems of modern vehicles, maximizing their potential through a dedicated control strategy. A control logic for the active management of RSD is introduced, designed to mitigate understeer or oversteer behavior in real time. System validation is initially performed in a simulation environment using standardized open-loop maneuvers in accordance with ISO protocols (Ramp Steer, Step Steer, Sine with Dwell). The controller is optimized using a Genetic Algorithm (GA) and subsequently validated through Driver-in-the-Loop (DiL) testing, conducted with a sample of 24 drivers at the dynamic simulator DriSMi of Politecnico di Milano. These tests aimed to assess both objective performance and subjective driver perception. Finally, the proposed system was integrated into a vehicle equipped with an Active Rear Steering system to explore the synergies between the two control systems and define their respective domains of effectiveness in lateral dynamics optimization.</div></div>","PeriodicalId":50615,"journal":{"name":"Control Engineering Practice","volume":"168 ","pages":"Article 106735"},"PeriodicalIF":4.6,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884374","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}
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