Robust Distributed Sensor Fault Detection and Diagnosis Within Formation Control of Multiagent Systems

This paper investigates the fault detection and diagnosis (FDD) problem for multiagent systems subject to sensor faults and disturbances. A distributed proportional integral derivative formation control protocol is constructed to achieve practical formation. A distributed FDD scheme, comprised of a fault detection module, a fault isolation module, and a fault estimation module, is designed within the formation control. For fault detection, the relationship between residuals and sensor faults in the multiagent systems is established such that each agent can detect the faults of all agents. For fault isolation, the distributed observer in each agent can determine whether the fault is in itself or its neighbors. Utilizing the absolute output information of each agent and the relative output information among neighboring agents, a distributed fault estimator is derived to provide the fault magnitude information. Simulation results are presented to demonstrate the effectiveness of the proposed FDD scheme.