This paper is devoted to the problem of distributed fault estimation for multiagent systems with nonlinear dynamic. For each agent, an augmented system can be obtained by treating the possible fault as a special system state. Utilizing the output of the neighbor agents, the nonlinear fault estimation observer can be constructed in one agent. The possible faults in both this agent and its neighbors can be estimated by the observer. By solving linear matrix inequalities, the parameter matrices of the observer can be computed. The region pole constraint is considered, which can improve the estimation performance. At last, both a numerical simulation and a network of four one-link flexible joint manipulator systems are considered to verify the proposed fault estimation approach.
Read full abstract