Multiple Faults Isolation for Hybrid Systems with Unknown Fault Pattern

ming yu, danwei wang, ming luo, and danhong zhang
Submission Type: 
Full Paper
phmc_10_071.pdf180.96 KBOctober 12, 2010 - 7:27am

This work is concerned with multiple faults isolation for hybrid systems based on Global Analytical Redundancy Relationships (GARRs) approach. GARRs are derived from the Hybrid Bond Graph (HBG) model of a hybrid system with a specified causality assignment procedure. In this article, multiple faults are considered in a complex hybrid system and these faults can develop during a mode when the faults are not detectable. Once a fault is detected, a fault candidates set is generated from mode dependent-fault signature matrix (MD-FSM) tables and a set of fault pattern hypothesis is created from the fault candidates set for further refinement. Fault isolation is carried out using a multiple nonlinear least square optimization (MNLSO) algorithm. The developed technique can deal with multiple faults with unknown pattern. The fault could be of incipient or abrupt nature. The simulation results show the effectiveness of the proposed method.

Publication Control Number: 
Submitted by: 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed