Neighborhood‐based distributed robust unknown input observer for fault estimation in nonlinear networked systems

Abstract

In this paper, a robust distributed fault estimation is addressed for a class of nonlinear networked systems with actuator faults where influences from process disturbances are minimized. In real-time monitoring systems, a substantial challenge is to find out the size and the shape of the occurred faults, and consequently, it is significantly more difficult in multi-agent systems. To consider this concern, in this paper, a distributed fault estimation approach has been proposed for multi-agent systems such that each agent employs an augmented system owing to a designated communication graph for estimating the fault and states both in this agent and in its neighbors. Meanwhile, each agent is equipped with an unknown input observer (UIO) to decouple the partial disturbances of the process as much as possible and to reduce the disturbances that cannot be decoupled. In the form of linear matrix inequalities (LMIs), sufficient conditions are provided to guarantee stability and to obtain the parameter matrices of the introduced observer. Finally, two numerical examples are provided to verify the performance of the introduced fault estimation scheme.