Distributed Interval Observer-based Fault Detection for a Class of Distributed Measurement Systems

This paper focuses on the fault detection problem for uncertain linear systems with distributed sensor networks. The fault detection algorithm is designed by using of distributed interval observer(DIO), which only uses the outputs of each agent and the communication from its neighbors to estimate the upper and lower bounds (ULBs) of the state of the whole system. The distributed interval estimation would be able, starting from coordinate transformation and vector rearrangement, to detect the fault by monitoring the unreasonable interval in each agent. The effectiveness of the proposed fault detection strategy has been demonstrated theoretically and numerically.