Safety Analysis Based On Bayesian Network For Cbtc System

Communication-based train control system adopted by Urban Transportation is a complex operation control system; it carries out its controlling functions integrated hardware and software. Therefore, there are some specified characters of CBTC, such as polymorphism for failure modes, independence of components, deviation of function and timing sequence. Meanwhile the safety of CBTC is affected by quality of software controlling, human factor and indirect factor, so that fault tree analysis and event tree analysis cannot satisfy.Compared with fault tree and event tree, Bayesian Network which has massy foundation of mathematics overcomes polymorphism, independence and timing problems, it presents dependence relationship between elements and components in probability graph, and the bi-direction reasoning can help us for prediction analysis in forward direction and diagnostic analysis in backward direction.According to the feature of CBTC, the advantages of Bayesian is adopted to carry out safety analysis for signaling system, integrated software problem, human factors and other uncertainty into analysis model.In terms of system function, Bayesian Network safety analysis Study Based on Scenario is raised in this paper with traditional functional analysis method and basic theory of Bayesian; firstly, we can obtain stimulus-response model for system deviation by goals and anti-goals of safety concern factors, and detailed scenarios for goals in specified scenario. Mapping the negative scenarios to Bayesian Network, we can carry out deduction analysis of supported scenarios for the condition possibility in major negative scenario. And then, we can calculate the possibilities for all of the nodes and make decisions with relative mitigating methods for anti-goals. At last this method is used to analyze safety separation which is a vital function in CBTC, the result shows that this function satisfies safety design requirement of CBTC.Bayesian Network safety analysis method based on scenario not only can use to analysis "safety separation" analysis but also is suitable for other functional safety analysis for CBTC...