多伺服系统的分布式自抗扰方法设计与仿真

Distributed Anti-disturbance Control and Simulation for Multi-servo Control Systems

在分布式多智能体系统中,最核心的问题就是如何设计出一个决策机制,通过基于通信拓扑结构可获取的本地信息来达到全局的一致性。提出了一种新型控制策略,将滑模变结构控制方法以及博弈论思想结合在一起。博弈论策略通过设计势函数以及梯度算法实现分布式系统的一致性;优化的滑模控制方法通过引入扩张状态观测器以及微分跟踪器实现对干扰的抑制以及对状态的估计。最后将该控制策略应用于二阶多伺服系统当中,建立实时仿真平台,通过半实物仿真证实了该方法的有效性。

In the distributed multi-agent systems,a major problem is how to design a decision-making architecture to achieve global consensus through available local information based on communication topology.This paper proposes a novel control strategy which combines sliding mode control with game theory.The game strategy realizes the consensus of the system by designing potential function and gradient algorithm.The optimized sliding mode control method realizes the ability of disturbance-rejection and state-estimation of the agents by introducing extended state observer and differential tracker.Finally,a real-time simulation platform is established and the control strategy is applied on the second-order multi-servo systems.The simulation results show the effectiveness of the method.