Application Research Of Bayesian Network In Probabilistic Safety Assessment Of CBTC System

As the nerve center of the train control system, the communication based train control(CBTC) system assuming heavy responsibilities of ensuring driving safety, improving operating efficiency and shorting the interval between trains. CBTC system that completes the train control functions by combining softwares and hardwares is a complex train control system and often goes wrongs which are hard to avoid. Though the softwares and hardwares of CBTC system own the function of safety guide in faults and dangerous side faults of system is considered to be low probability events, the consequence of dangerous side faults is very serious when dangerous side faults of system happen, in addition, security incidents of sysyem often result in serious consequences of equipment damage, property losses, casualties and so on, for these reasons, it is very important to make a safety assessment in CBTC system.Fauilt tree/event tree analysis(FTA/ETA) and bayesian network(BN) technology are combinied to estimate CBTC system at safety through analyzing the structures and functions of CBTC system. From the system failure perspective, CBTC system is systematically analyzed and estimated by taking into faults pattern characteristics for various factors of maintainability, polymorphism etc al. Through ensuring the potential risks, the causes of risk occurrences and the consequences of risk occurrences, the risk assessment models of CBTC system are established by BN, and then CBTC system are Estimated. The main research contents of article is as follows.(1) The structures of CBTC system are deeply analysised, its functions are devided hierar-chically. The system failure models of CBTC system are inducted and summarized from the system failure perspective, and then the failure fault tree model of CBTC system is established. On that basis, the fault tree model is taken into BN model and the availability of CBTC system is computed.(2) Given that the phenomena of maintainability and polymorphism of CBTC system failure patterns, the node evidences are input into BN model of failure system by BN technology who has powerful mathematical functions, then the failure datas of all nodes are calculated and the positive and negative reasoning are carried on to recognize the weak links of CBTC system. In addition, on the basis of BN model of failure system, the BN model of CBTC system based on fallback mode is built by the logic thought on analysis method of event tree, then the effect of the fallback mode on running safely of system is analyzed.(3) The risk assessment model of CBTC system based on BN is established through recognizing the potential risks, their consequences and probabilities of accidents caused by danger. The tolerable harzard rate(THR) of accidents probability caused by danger is calculated by the risk assessment model of CBTC system based on BN, then the CBTC system is judged to meet the safety requirements on and reach the security goals which are set,which provides help to evaluate scientifically the pobabilistic safety of CBTC system.