Consensus Control of Multi-Agent Systems Based on Adaptive Disturbance Observer

In this paper, a consensus problem of the directed connection topology with unknown external disturbances in multi-agent systems is considered. First, in order to approximate each agent's state and disturbance, an adaptive disturbance observer is presented. Next, based on the estimated value of the observer and related state information, a distributed consensus control protocol is designed. The conditions of global consensus control are determined by using the characteristics in real Jordan form of the Laplacian matrix. In particular, the proposed observer is completely decoupled from the consensus control. Then, the asymptotic convergence of the observer estimation error and the closed-loop stability of the proposed control protocol in time domain are analyzed under the framework of Lyapunov functions. Finally, simulations show the validity of the theoretical research.