检索结果-内蒙古大学图书馆

Degradation Tolerant Control Learning for Discrete-Time Affine Nonlinear Systems

IFAC-PapersOnLine 2023年第2期56卷 7734-7739页

作者： Soha KANSO Mayank Shekhar JHA Didier THEILLIOL CRAN UMR 7039 CNRS Université de Lorraine 54506 Vandoeuvre-lès-Nancy Cedex France

This paper develops a degradation tolerant optimal control in the framework of Reinforcement Learning (RL). Safety-critical and mission-critical systems require the development of new control designs that maintain system stability and performance specifications but also address incipient degradation. The aim of this work is to decelerate the speed of degradation by minimizing a cost function that includes the rate of evolution of degradation and the performance requirements. The controller is developed for discrete-time nonlinear systems affine in control, where the system's states are affected by a nonlinear degradation. Value iteration (VI) algorithm based approach is developed to find suitable approximations of both optimal control policy and optimal cost, while guaranteeing closed-loop stability and minimization of degradation rate. Offline model-based adaptive dynamic programming (ADP) algorithm is developed and implemented using actor-critic structure which involves training of both actor and critic neural networks (NN). After training the actor NN with the optimal policy, the NN is implemented in real time to generate the input of the system. Simulation example shows the efficiency and feasibility of the algorithm.

关键词： adaptive dynamic programming Reinforcement Learning Actor-Critic Structure Optimal Control Fault Tolerant Control Safety-Critical Systems Affine Nonlinear Systems

来源：评论

学校读者我要写书评

暂无评论

Learning-based adaptive Disturbance Rejection Control for Hypersonic Vehicle

Learning-based Adaptive Disturbance Rejection Control for Hy...

引用

第43届中国控制会议

作者： Xiao Han Bo Wang Lei Liu Huijin Fan Huazhong University of Science and Technology

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

This article investigates tracking control problem of hypersonic vehicles(HVs) with uncertain dynamics. To improve the control performance with no prior knowledge of modeling, an adaptive dynamic programming(ADP) based adaptive disturbance rejection control(ADRC) is proposed. First, a tracking differentiator(TD) and a neural extended state observer(NESO)are constructed to obtain differential command and resist uncertainty, respectively. Then, the error feedback design is adjusted by action-dependent heuristic dynamic programming(ADHDP) to approximate optimal control online. Especially, the proposed learning-based algorithm solves the parameter setting problem under uncertain dynamics by updating neural weights ***, simulation illustrates good tracking performance and strong adaptability compared with traditional ADRC.

关键词： Hypersonic Vehicles adaptive dynamic programming adaptive Disturbance Rejection Control Approximate Optimal Control

来源：评论

学校读者我要写书评

暂无评论

adaptive-Critic-Based Robust Control with Momentum Gradient Descent for Continuous-Time Nonlinear Systems

Adaptive-Critic-Based Robust Control with Momentum Gradient ...

引用

第43届中国控制会议

作者： Qinna Hu Ding Wang Ao Liu Faculty of Information Technology Beijing University of Technology Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing University of Technology Beijing Institute of Artificial Intelligence Beijing University of Technology Beijing Laboratory of Smart Environmental Protection Beijing University of Technology

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

In this paper, robust control problems are investigated for nonlinear continuous-time systems. A momentum-based gradient descent(GD) approach is developed to enhance the convergence performance of parameters in adaptive dynamic programming(ADP). By introducing the idea of momentum, the oscillation in the process of GD is alleviated and the selection of the learning rate becomes more flexible. Under the framework of ADP, the robust control problem is transformed into the optimal control problem by modifying the cost function. To avoid limitations of the initial admissible condition, an additional term is employed in the computation of the current gradient. Based on the online policy iteration algorithm, the momentum-based GD approach is constructed as an improved learning algorithm to optimize the critic network weights. Finally, a simulation is conducted to verify the effectiveness of the established learning strategy.

关键词： adaptive dynamic programming Momentum-based gradient descent Online policy iteration Optimal control Robust control

来源：评论

学校读者我要写书评

暂无评论

Decentralised adaptive learning-based control of robot manipulators with unknown parameters

引用

Journal of Automation and Intelligence 2025年

作者： Sveen, Emil Mühlbradt Zhou, Jing Ebbesen, Morten Kjeld Poursina, Mohammad Faculty of Engineering and Science University of Agder Jon Lilletuns vei 9 Agder Grimstad 4879 Norway

This paper studies motor joint control of a 4-degree-of-freedom (DoF) robotic manipulator using learning-based adaptive dynamic programming (ADP) approach. The manipulator's dynamics are modelled as an open-loop 4-link serial kinematic chain with 4 Degrees of Freedom (DoF). Decentralised optimal controllers are designed for each link using ADP approach based on a set of cost matrices and data collected from exploration trajectories. The proposed control strategy employs an off-line, off-policy iterative approach to derive four optimal control policies, one for each joint, under exploration strategies. The objective of the controller is to control the position of each joint. Simulation and experimental results show that four independent optimal controllers are found, each under similar exploration strategies, and the proposed ADP approach successfully yields optimal linear control policies despite the presence of these complexities. The experimental results conducted on the Quanser Qarm robotic platform demonstrate the effectiveness of the proposed ADP controllers in handling significant dynamic nonlinearities, such as actuation limitations, output saturation, and filter delays. © 2025 The Authors

关键词： 4-DoF adaptive dynamic programming Optimal control Robot manipulator Unknown dynamics

来源：评论

学校读者我要写书评

暂无评论

Decentralized Tracking Control of Interconnected Nonlinear Systems with Unmatched External Disturbances

Decentralized Tracking Control of Interconnected Nonlinear S...

引用

第43届中国控制会议

作者： Zeqiang Yuan Ding Wang Ao Liu Qinna Hu Zihang Zhou Faculty of Information Technology Beijing University of Technology Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing University of Technology Beijing Institute of Artificial Intelligence Beijing University of Technology Beijing Laboratory of Smart Environmental Protection Beijing University of Technology

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

In this paper, we develop a decentralized tracking control(DTC) strategy to stabilize a class of continuous-time nonlinear interconnected large-scale systems in the presence of unmatched external disturbances. This strategy is derived from an online learning approach of optimal control. Initially, the DTC problem is formulated to ensure both tracking control of isolated subsystems and the overall stability of the large-scale system. By combining the tracking error with the reference trajectory, some augmented systems are constructed and then the DTC design is transformed into an optimal control *** external disturbances, we formulated it as a two-player zero-sum game. Critic neural networks are utilized to solve the Hamilton-Jacobi-Isaacs equation. This approach allows for the estimation of Nash equilibrium solution that encompasses both the optimal control law and the worst-case disturbance law. Notably, a novel gradient descent strategy with momentum is established to tune the weights of the critic neural network for approximating the cost function better. Finally, an experimental simulation is provided to verify the effectiveness of the developed DTC scheme.

关键词： adaptive dynamic programming Decentralized control H_∞ control Reinforcement learning Tracking control

来源：评论

学校读者我要写书评

暂无评论

Optimal Control for Fractional-Order Nonlinear Systems Using Fractional-Order Online Policy Iteration

Optimal Control for Fractional-Order Nonlinear Systems Using...

引用

第43届中国控制会议

作者： Jie Kong Bo Zhao School of Systems Science Beijing Normal University

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

In this paper, a fractional-order online policy iteration(FOOPI)-based optimal control strategy is developed for a class of fractional-order nonlinear systems(FONSs). The fractional Taylor expansion and the property of Hadamard product are utilized to derive the fractional Hamilton-Jacobi-Bellman equation, which is solved by a FOOPI method consisting of policy evaluation and policy improvement. Through constructing a critic neural network(NN), whose weight vector is adjusted by the gradient descent algorithm, the cost function and the fractional control policy are obtained approximately. Then, the state trajectories of closed-loop FONSs and the critic NN weight vector approximation error dynamics are guaranteed to be uniformly ultimately bounded. Simulation results illustrate the effectiveness of the presented fractional optimal control scheme.

关键词： adaptive dynamic programming reinforcement learning optimal control neural networks fractional-order nonlinear systems

来源：评论

学校读者我要写书评

暂无评论

Integral Reinforcement Learning-Based Optimal Control for Nonzero-Sum Games of Multi-Player Input-Constrained Nonlinear Systems 7

Integral Reinforcement Learning-Based Optimal Control for No...

引用

7th International Conference on Robotics and Artificial Intelligence (ICRAI)

作者： Wu, Qiuye Zhao, Bo Liu, Derong Guangdong Univ Technol Guangzhou Guangdong Peoples R China Beijing Normal Univ Beijing Peoples R China

ISBN: (纸本)9781450385855

This paper investigates an integral reinforcement learning (IRL)based optimal control scheme to solve nonzero-sum games of multiplayer input-constrained nonlinear systems with unknown drift dynamics. The IRL method is introduced to obviate the identification procedure of the unknown drift dynamics. In order to achieve Nash equilibrium for each player, the simplified optimal control policy for each player is obtained by solving the coupled Hamilton-Jacobi equation with the critic neural network only. Thus, the computational resource is saved and the control structure is easy to realize. The closed-loop system is guaranteed to be uniformly ultimately bounded based on the Lyapunov stability analysis. Simulation results illustrate the effectiveness of the developed control scheme.

关键词： adaptive dynamic programming integral reinforcement learning neural networks nonzero-sum games input constraints

来源：评论

学校读者我要写书评

暂无评论

dynamic Event-Driven Decentralized Tracking Control of Input-Saturated Systems

Dynamic Event-Driven Decentralized Tracking Control of Input...

引用

第43届中国控制会议

作者： Wenqian Zheng Jianling Meng Xiong Yang School of Electrical and Information Engineering Tianjin University

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

This paper studies the decentralized tracking control(DTC) problem of asymmetrically input-saturated systems through a dynamic event-driven(ED) approach. First, the DTC problem is converted into the decentralized regulation problem of augmented systems. Then, it is further changed into a series of unsaturated optimal control problems of nominal subsystems with novel value functions. Meanwhile, the dynamic ED mechanisms and ED Hamilton-Jacobi-Bellman equations(ED-HJBEs) are presented. To solve the ED-HJBEs, only critic neural networks(NNs) are constructed with their weights updated in the adaptive dynamic programming framework. After that, the Lyapunov approach is applied to prove the uniform ultimate boundedness of tracking errors and critic NNs' weight estimation errors. Finally, simulations are provided to validate the proposed decentralized control scheme.

关键词： adaptive dynamic programming Asymmetric input saturation Decentralized tracking control Event-driven control

来源：评论

学校读者我要写书评

暂无评论

adaptive Optimal Control of Nonlinear Active Suspension Systems with Completely Unknown dynamics 33

Adaptive Optimal Control of Nonlinear Active Suspension Syst...

引用

33rd Chinese Control and Decision Conference (CCDC)

作者： Chen, Xin Huang, Yingbo Na, Jing Gao, Guanbin Zhao, Jun Kunming Univ Sci & Technol Fac Mech & Elect Engn Kunming 650500 Yunnan Peoples R China

ISBN: (纸本)9781665440899

Ride comfort, tire deflection and suspension stroke are generally conflicting in the active suspension control. To achieve a compromise among the three requirements, a novel online adaptive optimal control method is developed for the nonlinear active suspension systems subject to completely unknown dynamics. Based on the adaptive dynamic programming (ADP) algorithm, this paper proposes an identifier-critic structure with a dual neural network (NN) to realize the active suspension optimal control. To this end, a cost function concerning ride comfort, suspension stroke and control input is constructed. Then, an identifier NN with a novel adaptive law based on the parameter estimation error is used to approximate the unknown nonlinear dynamics. To obtain the optimal solution, a critic NN with the proposed adaptive law is suggested to design the optimal controller. Simulation results are provided to verify the effectiveness of the proposed method.

关键词： Active suspension systems optimal control active suspension control adaptive dynamic programming

来源：评论

学校读者我要写书评

暂无评论

Event-Driven Optimal Regulation with SNAC Framework for Constrained Nonlinear Plants 40

Event-Driven Optimal Regulation with SNAC Framework for Cons...

引用

40th Chinese Control Conference (CCC)

作者： Hu, Lingzhi Wang, Ding Zhao, Mingming Ren, Jin Beijing Univ Technol Fac Informat Technol Beijing 100124 Peoples R China Bejing Univ Technol Beijing Key Lab Computat Intelligence & Intellige Beijing 100124 Peoples R China Beijing Univ Technol Beijing Inst Artificial Intelligence Beijing 100124 Peoples R China

ISBN: (纸本)9789881563804

In order to reduce the computational burden effectively and ensure that the system has good control performance, an event-driven near-optimal control algorithm is developed for discrete-time nonlinear systems with input constraints. A single network adaptive critic algorithm is adopted to address the constrained optimal control problem. First, the optimal cost function and the optimal control strategy are obtained theoretically through the iterative algorithm. Then, an appropriate triggering threshold is established, so that the system satisfies the event-driven mechanism. In addition, the action network of classical adaptive dynamic programming is omitted. The derivative of the approximate cost function is the output of the critic network and then is used to obtain the approximate optimal control law. Furthermore, the convergence analysis of the iterative algorithm is proved. Finally, a simulation example is conducted to verify the effectiveness of the proposed scheme.

关键词： adaptive dynamic programming Control Constraints Discrete-Time Systems Event-Driven Control Neural Networks Single Network adaptive Critic

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：