Fault diagnosis for time-varying systems with multiplicative noises over sensor networks subject to Round-Robin protocol

Abstract

In this paper, the fault diagnosis problem is investigated for time-varying systems with multiplicative noises over sensor networks. The system model is subject to state- and control-dependent noises, and the Round-Robin protocol is applied in sensor networks to prevent data collisions. By means of the least-squares approach, a series of fault estimation filters are designed for the addressed system, and each filter can estimate the augmented state which includes system state and a fault signal. According to residual signals, several evaluation functions are derived and one of them is selected for the purpose of fault detection. By employing the residual matching method, the filter with least evaluation function can be used to isolate and estimate the occurred fault. A numerical example is provided to demonstrate the effectiveness of our proposed fault diagnosis scheme.