Plug-and-Play Robust Distributed Fault Estimation for Interconnected Systems

Abstract

In this paper, a robust distributed fault estimation method is studied for interconnected systems with process and measurement disturbances. Each subsystem is monitored by an unknown input observer (UIO) based fault estimator, using local output measurement and estimated states of neighbors. The existence conditions of the distributed fault estimators (DFEs) are given in a decentralized fashion. Then, robust DFEs are designed to constrain the influence of the disturbances on the fault estimation errors, of which the global design conditions are transformed into a linear matrix inequality (LMI). Particularly, if the completely decoupling conditions are satisfied, the DFEs have an asymptotic estimation feature, and can be constructed using only local and interconnected information. Furthermore, the proposed robust DFEs are allowed for plug-and-play operations, and the corresponding conditions are analyzed in detail. Simulation studied on two cases is conducted to illustrate the effectiveness of the proposed method.