{"title":"基于强化学习的受扰多代理系统事件触发约束遏制控制","authors":"Daocheng Tang;Ning Pang;Xin Wang","doi":"10.1109/TSIPN.2024.3487422","DOIUrl":null,"url":null,"abstract":"This article investigates the full-state-constrained optimal containment control problem of perturbed nonlinear multiagent systems (MASs). Initially, to balance control accuracy and cost while maintaining the states of MASs within confined regions, an enhanced constrained optimized backstepping (OB) framework is first developed for the multiagent control scenario by adopting an identifier-actor-critic-based reinforcement learning (RL) algorithm, where a novel performance index based on the barrier Lyapunov function (BLF) is integrated into the classic OB framework. Then, to enhance the robustness of the systems, the proposed framework employs disturbance observers to mitigate the effects of unknown external disturbances. Moreover, sufficient conditions are established to ensure that systems maintain stability and expected performance under denial-of-service (DoS) attacks. Subsequently, the controller implements a novel dynamic event-triggered mechanism (DETM), adaptively adjusting the triggering conditions by the estimated neural network (NN) weights in the proposed framework for substantial communication burden reduction. Finally, the stability of the systems is demonstrated using the Lyapunov theory, and a simulation example confirms the feasibility of the proposed scheme.","PeriodicalId":56268,"journal":{"name":"IEEE Transactions on Signal and Information Processing over Networks","volume":"10 ","pages":"820-832"},"PeriodicalIF":3.0000,"publicationDate":"2024-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Reinforcement Learning-Based Event-Triggered Constrained Containment Control for Perturbed Multiagent Systems\",\"authors\":\"Daocheng Tang;Ning Pang;Xin Wang\",\"doi\":\"10.1109/TSIPN.2024.3487422\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This article investigates the full-state-constrained optimal containment control problem of perturbed nonlinear multiagent systems (MASs). Initially, to balance control accuracy and cost while maintaining the states of MASs within confined regions, an enhanced constrained optimized backstepping (OB) framework is first developed for the multiagent control scenario by adopting an identifier-actor-critic-based reinforcement learning (RL) algorithm, where a novel performance index based on the barrier Lyapunov function (BLF) is integrated into the classic OB framework. Then, to enhance the robustness of the systems, the proposed framework employs disturbance observers to mitigate the effects of unknown external disturbances. Moreover, sufficient conditions are established to ensure that systems maintain stability and expected performance under denial-of-service (DoS) attacks. Subsequently, the controller implements a novel dynamic event-triggered mechanism (DETM), adaptively adjusting the triggering conditions by the estimated neural network (NN) weights in the proposed framework for substantial communication burden reduction. Finally, the stability of the systems is demonstrated using the Lyapunov theory, and a simulation example confirms the feasibility of the proposed scheme.\",\"PeriodicalId\":56268,\"journal\":{\"name\":\"IEEE Transactions on Signal and Information Processing over Networks\",\"volume\":\"10 \",\"pages\":\"820-832\"},\"PeriodicalIF\":3.0000,\"publicationDate\":\"2024-11-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Transactions on Signal and Information Processing over Networks\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10747699/\",\"RegionNum\":3,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Signal and Information Processing over Networks","FirstCategoryId":"94","ListUrlMain":"https://ieeexplore.ieee.org/document/10747699/","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
摘要
本文研究了扰动非线性多代理系统(MAS)的全状态约束优化控制问题。首先,为了在将 MAS 的状态保持在受限区域内的同时平衡控制精度和成本,本文针对多代理控制场景,通过采用基于识别器-代理-批判的强化学习(RL)算法,开发了增强型受限优化反步态(OB)框架,并将基于障碍李亚普诺夫函数(BLF)的新型性能指标集成到经典的 OB 框架中。然后,为了增强系统的鲁棒性,所提出的框架采用了干扰观测器来减轻未知外部干扰的影响。此外,还建立了充分条件,以确保系统在拒绝服务(DoS)攻击下保持稳定和预期性能。随后,控制器实施了一种新颖的动态事件触发机制(DETM),通过估计拟议框架中的神经网络(NN)权重自适应地调整触发条件,从而大大减轻了通信负担。最后,利用 Lyapunov 理论证明了系统的稳定性,一个仿真实例证实了所提方案的可行性。
Reinforcement Learning-Based Event-Triggered Constrained Containment Control for Perturbed Multiagent Systems
This article investigates the full-state-constrained optimal containment control problem of perturbed nonlinear multiagent systems (MASs). Initially, to balance control accuracy and cost while maintaining the states of MASs within confined regions, an enhanced constrained optimized backstepping (OB) framework is first developed for the multiagent control scenario by adopting an identifier-actor-critic-based reinforcement learning (RL) algorithm, where a novel performance index based on the barrier Lyapunov function (BLF) is integrated into the classic OB framework. Then, to enhance the robustness of the systems, the proposed framework employs disturbance observers to mitigate the effects of unknown external disturbances. Moreover, sufficient conditions are established to ensure that systems maintain stability and expected performance under denial-of-service (DoS) attacks. Subsequently, the controller implements a novel dynamic event-triggered mechanism (DETM), adaptively adjusting the triggering conditions by the estimated neural network (NN) weights in the proposed framework for substantial communication burden reduction. Finally, the stability of the systems is demonstrated using the Lyapunov theory, and a simulation example confirms the feasibility of the proposed scheme.
期刊介绍:
The IEEE Transactions on Signal and Information Processing over Networks publishes high-quality papers that extend the classical notions of processing of signals defined over vector spaces (e.g. time and space) to processing of signals and information (data) defined over networks, potentially dynamically varying. In signal processing over networks, the topology of the network may define structural relationships in the data, or may constrain processing of the data. Topics include distributed algorithms for filtering, detection, estimation, adaptation and learning, model selection, data fusion, and diffusion or evolution of information over such networks, and applications of distributed signal processing.