This paper presents a design of robust consensus for homogeneous leader-follower multi-agent systems (MAS). Each agent of MAS is described by a linear time-invariant dynamic model subject to parametric uncertainty. The agents are interconnected through a static interconnection matrix over an undirected graph to cooperate and share information with their neighbours. The consensus design of MAS can be transformed to stability analysis by using decomposition technique. We apply Lyapunov theorem to derive the sufficient condition to ensure the consensus of all independent subsystems. In addition, we design a robust distributed state feedback gain based on linear quadratic regulator (LQR) setting. Controller gain is computed via solving a linear matrix inequality. As a result, we provide a robust design procedure of a cooperative LQR control to achieve consensus objective and maximize the admissible bound of the uncertainty. Finally, we give numerical examples to illustrate the effectiveness of the proposed consensus design. The results show that the response for MAS in presence of uncertainty using robust consensus design follows the response of the leader and is better than that of the existing nominal consensus design.
All Science Journal Classification (ASJC) codes