Train Control System Reliability Real-Time Analysis Method Based On Resilience Effect

The operation safety of high-speed train is very important. Train control system with critical safety characteristics is one of the key systems for safe operation of high-speed train. The train control system with complex structure is sensitive to internal factors and external environment in the operational process. If any small faults or safety problems of the system can not be diagnosed in real time and handled effectively, it is like to trigger a chain reaction or disastrous consequences. So the operational reliability of train control system should be studied intensively. Resilience, which is accorded to the resilient characteristics of the system, is a system force that will be adopted in the case of system failure or maintenance. It begins to function when the reliability of system becomes low. By this way, the operation reliability of system can be improved. In this dissertation, according to the characteristics of the train control system, the resilience concept of train control system is defined and the feature of resilient process is analyzed. Then a set of reliability analysis methods is designed and applied to the resilient process of train control system. According to historical failure data, the reliability analysis of different levels of train control system is carried out by using complex network and Bayesian network methods. Finally, according to the results of reliability analysis, function improvement and maintenance strategy are proposed based on resilience.The main innovation of this dissertation is as follows:(1) Resilience which is applied to train control system is defined and the characteristics of different stages of resilient process are described. On this basis, a set of methods for reliability of resilient process of train control system is designed. (2) Based on complex network theory, the information exchange process between subsystems, subsystem and train is abstracted to a network model. Then the reliability of system is analyzed separately in the cases of normal, single node failure and edges failure. According to resilience evaluation index, the fault response capability of subsystem is evaluated. By above research, key equipment and weak links, which make effects on the reliability of train control system, are determined. The hardware and function reliability of subsystem are analyzed by using Bayesian network method. The discrete time Bayesian network of on-board equipment is established by priori knowledge. Then in Matlab simulation environment, operation reliability change of subsystem is calculated by using Bayesian network toolbox. Finally, the core function reliability of safety computer is evaluated by using static Bayesian network. (3) Based on the reliability analysis of train control system, specific function improvement is proposed for train control system in operation stage, and maintenance strategy is formulated after system failure. Specific analysis on the function improvement and maintenance plan of on-board equipment is conducted.A set of reliability analysis methods is designed and a scheme is put forward to improve the reliability of system in this dissertation. And the historical failure data is used for analysis. These methods are applied in the train control system, which provides effective theory foundation and technical guidance for the real-time reliability evaluation of high-speed train control system.