Fault diagnosis and prediction of hybrid system based on particle filter algorithm

Abstract

Hybrid system consists of discrete and continuous state. Fault diagnosis and prediction of hybrid system is a new and critical issue in hybrid system study. Utilizing the advantage of particle filter, which can estimate the discrete and continuous states simultaneously, a new method for fault diagnosing and predicting of hybrid system is proposed. After giving the design steps of particle filter based fault diagnosis and prediction, we discuss the application of the method to the mobile robot fault diagnosis and prediction system. The simulation results show the effectiveness of the proposed approach.