Adaptive Tracking Control Under Switched and Fixed Topologies With a Leader’s Unknow and Bounded Inputs for Linear Multi-Agent Systems

Abstract

This document tackles the issue of distributed adaptive tracking control for generic linear multi-agent systems. Specifically, the focus is on linear systems with non-zero control inputs from the leader, which are not directly accessible to the follower. Unlike existing literature limited to fixed topologies, this study tackles the challenge of switching topologies. A distributed discontinuous adaptive tracking controller uses Riccati inequalities and based on the state errors between neighboring agents. This controller adaptively adjusts the coupling weights between adjacent agents and ensures that switching communication topologies does not affect follower agent communication and remains undirected, allowing the leader to reach all followers via a directed path in the joint graph. This paper also examines stable multi-agent systems using fixed and dynamic network configurations. A communication graph in which the leader is bound by their control inputs is shown to be able to demonstrate such behavior, provided that the graph is jointly connected. A description of the followers’ states will effectively approach and converge on the leader’s state. An additional simulation experiment confirms its effectiveness and robustness of suggested method.