检索结果-内蒙古大学图书馆

40th Chinese Control Conference (CCC)

作者： Zhao, Huiling Wang, Ding Zhao, Mingming Ren, Jin Beijing Univ Technol Fac Informat Technol Beijing 100124 Peoples R China Beijing 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 implement the tracking control design towards concentrations of dissolved oxygen and nitrate nitrogen in wastewater treatment, a data-based iterative dual heuristic dynamic programming (DHP) method is established in this paper. First, the optimal tracking control problem is transformed into the optimal regulation problem. Then, the iterative adaptive dynamic programming strategy is introduced to solve the optimal control problem of the new system with the discount factor in the cost function, and the convergence analysis of the iterative algorithm is given. Furthermore, a new training approach is used to adjust the weights of the action and critic networks. Finally, the experimental simulation for a wastewater treatment plant is presented to verify the applicability of the proposed iterative DHP optimal tracking control method.

关键词： adaptive dynamic programming Discount factor Neural networks Optimal tracking control Wastewater treatment.

来源：评论

学校读者我要写书评

暂无评论

adaptive dynamic programming-based State Quantized Networked Control System without Value and/or Policy Iterations

Adaptive Dynamic Programming-based State Quantized Networked...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Zhao, Qiming Xu, Hao Jagannathan, S. Missouri Univ S&T Dept Elect & Comp Engn Rolla MO 65409 USA

ISBN: (纸本)9781467314909

In this paper, the Bellman equation is used to solve the stochastic optimal control of unknown linear discrete-time system with communication imperfections including random delays, packet losses and quantization. A dynamic quantizer for the sensor measurements is proposed which essentially provides system states to the controller. To eliminate the effect of the quantization error, the dynamics of the quantization error bound and an update law for tuning its range are derived. Subsequently, by using adaptive dynamic programming technique, the infinite horizon optimal regulation of the uncertain NCS is solved in a forward-in-time manner without using value and/or policy iterations by using Q-function and reinforcement learning. The asymptotic stability of the closed-loop system is verified by standard Lyapunov stability theory. Finally, the effectiveness of the proposed method is verified by simulation results.

关键词： Networked Control System Quantization adaptive dynamic programming Optimal Control

来源：评论

学校读者我要写书评

暂无评论

Generalized Policy Iteration-based Reinforcement Learning Algorithm for Optimal Control of Unknown Discrete-time Systems 33

Generalized Policy Iteration-based Reinforcement Learning Al...

引用

33rd Chinese Control and Decision Conference (CCDC)

作者： Lin, Mingduo Zhao, Bo Liu, Derong Liu, Xi Luo, Fangchao Guangdong Univ Technol Sch Automat Guangzhou 510006 Peoples R China Beijing Normal Univ Sch Syst Sci Beijing 100875 Peoples R China

ISBN: (纸本)9781665440899

This paper presents a novel generalized policy iteration-based reinforcement learning (RL) algorithm to deal with infinite-horizon optimal control problems of nonlinear discrete-time systems with completely unknown dynamics. In the present iterative algorithm, two iteration procedures are utilized to obtain the iterative Q-function and the iterative control policy. Furthermore, the iterative Q-function is obtained by the temporal difference learning and the policy gradient method is utilized to directly optimize the iterative control policy. Then, the convergence and optimality analysis of the generalized policy iteration-based RL algorithm are provided. To implement this algorithm, two neural networks, including a critic network and an action network, are used to approximate the iterative Q-function and the iterative control policy. Finally, a numerical simulation example is provided to illustrate the effectiveness of the proposed control method.

关键词： adaptive dynamic programming Reinforcement learning Generalized policy iteration Neural networks Optimal control Unknown Discrete-time Systems

来源：评论

学校读者我要写书评

暂无评论

Fault-tolerant control for nonlinear offshore steel jacket platforms based on reinforcement learning

引用

OCEAN ENGINEERING 2022年 246卷 110247-110247页

作者： Ziaei, Amin Kharrati, Hamed Rahimi, Afshin Univ Tabriz Fac Elect & Comp Engn Tabriz Iran Univ Windsor Dept Mech Automot & Mat Engn Windsor ON N9B 3P4 Canada

In this article, reinforcement learning based fault-tolerant control for the nonlinear offshore platform is studied. First, the dynamic model of the nonlinear steel-jacket offshore platform with active mass damper and actuator fault is investigated then an affine nonlinear representation for the offshore platform is obtained. By estimating actuator fault and irregular wave force and inserting their estimation into the cost function, the disturbance control and fault-tolerant control problem (FTC) are converted into optimal control problem. Online policy iteration is used to minimize the performance index. Finally, simulation results show the effectiveness of our method.

关键词： Nonlinear steel-jacket offshore platforms Fault-tolerant control Fault observer Reinforcement learning adaptive dynamic programming Disturbance observer

来源：评论

学校读者我要写书评

暂无评论

Neural-Network-Based Self-Learning Disturbance Rejection Design for Continuous-Time Nonlinear Constrained Systems 40

Neural-Network-Based Self-Learning Disturbance Rejection Des...

引用

40th Chinese Control Conference (CCC)

作者： Li, Menghua Wang, Ding Qiao, Junfei Xu, Xin Beijing Univ Technol Fac Informat Technol Beijing 100124 Peoples R China Beijing 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 this paper, we adopt a neural-network-based adaptive dynamic programming (ADP) method to solve the disturbance rejection problem for continuous-time nonlinear systems with control constraints. First, we define a suitable function and then derive the optimal control law with control constraints and the worst disturbance law. Besides, the constrained Hamilton-Jacobi-Isaacs equation for continuous-time nonlinear constrained systems is derived. Then, only one critic neural network is used to approximate the optimal cost function. Consequently, the approximate optimal control law and the approximate worst disturbance law are obtained. Additionally, a new updating rule is developed in the process of neural critic learning. Finally, the simulation results show that the self-learning optimal control is realized and the effectiveness of the disturbance rejection is verified by using the neural-network-based ADP method.

关键词： Optimal control Nonlinear constrained systems adaptive dynamic programming Neural networks Self-learning disturbance rejection H-infinity control

来源：评论

学校读者我要写书评

暂无评论

Observer-critic-based Event-triggered Decentralized Optimal Control of Modular Robot Manipulators 40

Observer-critic-based Event-triggered Decentralized Optimal ...

引用

40th Chinese Control Conference (CCC)

作者： Pan, Qiang Ma, Bing Dong, Bo Ren, Xiaolin Li, Yuanchun Changchun Univ Technol Dept Control Sci & Engn Changchun 130000 Peoples R China

ISBN: (纸本)9789881563804

This paper presents a novel observer-critic-based event-triggered decentralized optimal tracking control of modular robot manipulators (MRMs). Based on the dynamic model of the MRM subsystem utilizing joint torque feedback (JTF) technique, the performance index function is constructed by integrating the designed adaptive observer and the tracking error fusion function including position error and velocity error. On the basis of adaptive dynamic programming (ADP) algorithm and the event-triggered mechanism, the decentralized optimal tracking control problem is settled by solving the Hamiltonian-Jacobi-Bellman (HJB) equation via constructing the critic neural network (NN). The closed-loop MRM system is proved to be asymptotically stable by using the Lyapunov stability theorem. Finally, experimental results illustrate the effectiveness of the developed control method.

关键词： Modular robot manipulators adaptive dynamic programming event-triggered mechanism decentralized optimal control neural network

来源：评论

学校读者我要写书评

暂无评论

An Intelligent and Secure Control Approach for Nonlinear Systems under Attacks

An Intelligent and Secure Control Approach for Nonlinear Sys...

引用

IEEE Symposium Series on Computational Intelligence (IEEE SSCI)

作者： Zhong, Xiangnan Ni, Zhen Florida Atlantic Univ Dept Elect Engn & Comp Sci Boca Raton FL 33431 USA

ISBN: (纸本)9781728190488

This paper designs an intelligent and secure control approach based on adaptive dynamic programming for a class of nonlinear systems under the actuator attacks. The designed method can monitor and identify the attacks by an established state estimator based detector. When an attack is triggered, the control process automatically switches to a game-theoretical architecture for attack mitigation. The intelligent learning process is developed for both nominal and attack conditions without the requirement of system dynamics and attack information. Neural network techniques are applied to implement the proposed method with two critic networks and the control signals are calculated accordingly. Therefore, the designed intelligent control method can reduce the computation complexity. Simulation studies and results demonstrate the necessary of attack detection and mitigation during the learning process, and also verify the effectiveness of the developed method.

关键词： adaptive dynamic programming reinforcement learning attack mitigation intelligent control attack detection and identification

来源：评论

学校读者我要写书评

暂无评论

adaptive neural control of PEMFC system based on data-driven and reinforcement learning approaches

引用

CONTROL ENGINEERING PRACTICE 2022年第0期120卷 105022-105022页

作者： Lin-Kwong-Chon, Christophe Damour, Cedric Benne, Michel Kadjo, Jean-Jacques Amangoua Grondin-Perez, Brigitte Univ Reunion Isl LE2P EnergyLab EA 4079 15 Ave Rene CassinBP 7151 F-97715 St Denis Reunion France

Proton exchange membrane fuel cell systems are being increasingly put forward as hydrogen energy carrier converters. Recent advancements in reliability strategies have been stimulated through maintaining a healthy operating condition of the system and covering plant faults. However, it is observed that occurrence or even the mitigation of these faults cause multilateral effects that can potentially destabilize the normal operation of the system. In the active fault tolerant control strategy, two modules are designed to fault management. The diagnostic module identifies the apparent fault and identifies the corrective commands, then the re-design module adapts the controller to dynamic system changes. In order to improve the generic characteristics of the re-design module, this paper presents a data-driven neural controller capable to automatically adapt to system health states. The developed approach comes from the machine learning class and combines adaptive dynamic programming, deep echo-state neural network models and fuzzy logic learning. The proposed controller is evaluated under occurrence of channels flooding and membrane drying faults, but also actuators and water purging disturbances. Simulation and experimental results show the effectiveness of the proposed data-driven approach without prior neural model training, while guaranteeing the stability and learning convergence of the adaptive controller.

关键词： Proton exchange membrane fuel cell Active fault tolerance adaptive dynamic programming Reinforcement learning Deep echo state network

来源：评论

学校读者我要写书评

暂无评论

Zero-sum game-based security control of unknown nonlinear Markov jump systems under false data injection attacks

引用

INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL 2022年第0期

作者： Gao, Xiaobin Deng, Feiqi Zeng, Pengyu South China Univ Technol Sch Automat Sci & Engn Guangzhou 510640 Guangdong Peoples R China

In order to solve the security control problem of discrete-time unknown Markov jump systems (MJSs) subject to false data injection attacks, a new zero-sum game-based control method is developed in this paper by using model-free adaptive dynamic programming algorithm. The controller and the attacker are regarded as two players. An optimal cost function of MJSs is firstly introduced based on the evolution of the Markov chain. Then, a critic-action-attack framework is established to approximate the cost function, control input and false data, respectively. Additionally, the boundedness of the neural network (NN) weight estimate errors is guaranteed via Lyapunov theory. In the end, the validity of the proposed approach is verified by simulation of a single link robot arm.

关键词： adaptive dynamic programming cyber attacks Markov jump systems security control zero-sum game

来源：评论

学校读者我要写书评

暂无评论

adaptive optimal output regulation of linear discrete-time systems based on event-triggered output-feedback

引用

AUTOMATICA 2022年 137卷 110103-110103页

作者： Zhao, Fuyu Gao, Weinan Liu, Tengfei Jiang, Zhong-Ping Northeastern Univ State Key Lab Synthet Automat Proc Ind Shenyang 110004 Peoples R China Florida Inst Technol Coll Engn & Sci Dept Mech & Civil Engn Melbourne FL 32901 USA NYU 6 Metrotech Ctr Dept Elect & Comp Engn Brooklyn NY 11201 USA

This paper presents novel event-triggered control approaches to solve the adaptive optimal output regulation problem for a class of linear discrete-time systems. Different from most existing research on output regulation problems, the developed adaptive optimal control approaches are based on (1) output-feedback instead of full-state or partial-state feedback, (2) adaptive dynamic programming (ADP) which provides approximate solutions of the optimal control problem without requiring the precise knowledge of the plant dynamics, and (3) an event-triggering mechanism that reduces the communication between the controller and the plant. It is shown that the system in closed-loop with the developed controllers is asymptotically stable at an equilibrium of interest, and the tracking errors asymptotically converge to zero. Moreover, the suboptimality of the closed-loop system is directly determined by the relative threshold, which is a ratio between the triggering threshold and the actual state. A numerical simulation example is employed to verify the effectiveness of the proposed methodologies. (C) 2021 Elsevier Ltd. All rights reserved.

关键词： adaptive dynamic programming Output regulation Event-triggered control

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：