Consensus Control of Multi-agent Systems Based on Fault Estimation

Abstract

This paper provides a consensus control scheme with a novel fault compensation for multi-agent systems with respect to occurred faults. To this end, a novel estimator is first developed via injecting the noise bias to satisfy the unbiasedness requirement. Benefiting from the estimated fault, a consensus controller with fault compensation is constructed to realize the desired l2-l∞ performance while guaranteing fundamental consensus. According to the technology of variance analysis combined with the method of Lyapunov stability, sufficient conditions are strictly derived to ensure the predetermined performance and the applicable gains of both the estimator and the controller are attained via a recursive formula and linear matrix inequalities. Finally, the applicability is tested by a simulation example.