Fully Distributed Adaptive State Estimation and Consensus Control of Multi-agent Systems: A Reduced-order Observer-based Approach

This paper focuses on the integrated design of adaptive state estimation and consensus control for multi-agent systems. To improve the consensus performance of the multi-agent systems in the presence of some agents’ sensor failures, the cooperative state estimation model is introduced, and fully distributed adaptive output tracking estimators and reduced-order state observers are designed based on this model. The consensus control protocol, relying on only the self-estimated states, is proposed and enables the achievement of consensus of the multi-agent system under directed graphs. The effectiveness of the algorithm is theoretically analyzed and validated through simulation.