Distributed Fault Estimation Over Sensor Networks With Randomly Switching Topologies Under Binary Encoding Scheme

Abstract

In this article, a distributed fault estimator is designed for nonlinear systems over sensor networks subject to randomly switching topologies under binary encoding scheme. Binary encoding scheme, which enhances security and decreases energy occupation of network communication, is utilized to schedule innovation transmissions among sensor nodes. In order to characterize actual transmission situation, random bit errors are considered whose influence is expressed by an additive stochastic noise. Moreover, a Bernoulli-distributed random variable is put into use to characterize the randomly switching topologies. This article aims to design distributed fault estimators such that the estimation error dynamics is exponentially ultimately bounded in mean square. A sufficient condition is set up for designing the estimator by means of statistical properties analysis and matrix inequality technique. Gain parameters of the distributed estimator are determined via solving a group of matrix inequalities. The simulation results validate that the proposed distributed fault estimation approach is effective.