多体线性系统一致性分布观测控制器设计

Distributed Consensus Observer and Controller Design for Multi-Agent Linear Systems

现有跟踪自治线性领导者系统的一致性算法在领导者带有控制输入的情况下会有较大的跟踪误差,因此设计了一种在切换网络条件下,针对领导者、跟随者均为一般线性系统的多体一致性分布式观测器与控制器。设计使用邻近节点的局部信息状态,通过构造分布式观测器来重构领导者系统的参考状态与控制激励,并使用各跟随节点的观测值构造控制器来完成对领导者的一致性控制。此外,轮式机器人编队控制实例说明了该算法在跟随节点需要保持可变相对状态偏差条件下的应用方法,仿真结果验证了算法的有效性。

The present consensus algorithm for tracking autonomous linear leader systems will cause a largetracking error in the case that the leader system has control inputs, so distributed observers and controllers formulti-agent linear systems are designed for the general linear leader-following system under the switchingnetwork. The design uses the local information state from neighbor nodes, reconstructs the reference states andcontrol inputs of the leader system by using a distributed observer and constructs a controller by using theobserved value to solve the leader tracking consensus problem. In addition, a wheeled robot formation exampleillustrates the application method of the algorithm when the follower nodes need to keep a variable relativestate deviation. The simulation results prove the validity of the algorithm.