带有输出约束条件的随机多智能体系统容错控制

Fault-tolerant control for stochastic multi-agent systems with output constraints

中文关键词  非严格反馈形式  容错控制  多智能体系统  输出约束  随机系统

英文关键词  nonstrict-feedback form  fault-tolerant control  multi-agent systems  output constraints  stochastic systems

中文摘要

      在有向通讯拓扑图下, 针对一类具有输出约束和执行器偏差增益故障的非严格反馈随机多智能体系统, 提 出一种自适应神经网络容错控制设计方案. 采用神经网络逼近未知非线性函数, 构造障碍李雅普诺夫函数处理系统 的输出约束问题, 以反步法和动态面技术为框架, 结合Nussbaum函数设计自适应神经网络容错控制方法. 基于李雅 普诺夫稳定性理论, 证明所有跟随者输出与领导者输出达到一致, 闭环系统的所有信号依概率半全局一致最终有界 且系统输出限制在给定紧集内. 论文最后通过仿真实验验证所给出控制方案的有效性.

英文摘要

      In this paper, the adaptive neural network fault-tolerant control strategy is proposed for a class of nonstrictfeedback stochastic multi-agent systems with output constraints and actuator faults under a directed communication topology. Neural networks are utilized to approximate unknown nonlinear functions. Furthermore, the barrier Lyapunov function is employed to deal with the problem of output constraints. Combining backstepping method, dynamic surface control technique and Nussbaum function, an adaptive neural network fault-tolerant control method is proposed. Based on Lyapunov stability theory, it is proved that all followers’ outputs can be consistent with the leader’s output. All signals in the closedloop systems are semiglobally uniform ultimate bounded in probability and the output of systems can be limited within the given compact set. Finally, the effectiveness of the proposed control scheme is verified through numerical simulation.