Study of fault diagnosis method based on fuzzy Bayesian network and application in CTCS-3 train control system

Abstract

The fault diagnosis approach based-on Bayesian network is frequently used in fault diagnosis field, has better completeness and explanation facility, but it has some disadvantages due to lack of the quantitative data information, and it is difficult to applied in complicated system, China train control system lever-3 (CTCS-3) is a complicated system with high security requirement, so a Bayesian fuzzy inference nets real-time internal fault diagnostic system for CTCS-3 train control system is proposed. The membership functions and symptom-fault mapping relationship for CTCS-3 fault diagnosis system are obtained from pre-measured experimental data as well as experts' diagnostic experience/knowledge to distinguish the effect of true fault from various factors. The cores dangerous of train control system are supervision and protect against exceedance of safe speed distance and driver exceeds safe speed distance, according the two safety links, first of all, set up fault tree; and then convert to Bayesian network, finally the fuzzy Bayesian network diagnosis arithmetic of fault diagnosis system with accuracy is designed and presented. The validity and effectiveness of the proposed approach is witnessed clearly from the testing results obtained. In the last part of the paper, the fault diagnosis system of CTCS-3 is established by using the fuzzy inference algorithm.