Functional Safety Analysis Of Train Controlling Of RBC Scenes Based On UML Extension Mechanism

With the development of high-speed railway technology in China and the promulgation of high-speed railway management rules and regulations,China’s requirements for CTCS-3(China Train Control System at level 3)have also gradually improved.CTCS-3 is a hybrid system with both continuous and discrete information interaction mechanisms,when RBC interacts with other equipment,any problems related to driving permission,route information,etc.may affect the normal operation of high-speed trains.Therefore,a safety analysis of the train controlling of RBC scenes is required to ensure the subsequent safety design and evaluation of the system.However,the traditional security assurance technology has slightly deficiencies in the security analysis of the train control system,and the formal analysis of the train control system’s safety has become the main trend of research.This thesis takes the train controlling of RBC scenes in the CTCS-3 as the research object,and proposes a functional safety analysis method for the train controlling of RBC scenes based on the UML(Unified Modeling Language)extension mechanism.Firstly,aiming at the hybrid characteristics of the CTCS-3,an object-oriented method is used to analyze the train controlling of RBC scenes.According to the security characteristics modeling method of the UML extension mechanism,a UML class diagram and a state transition diagram model and its fault extension model are constructed to describe the state transition and continuous dynamic process of the system.Then,according to the requirements of the train control system,a hybrid profile with security characteristics is constructed to accurately describe the hybrid characteristics of the train control system.Secondly,selecting CPN(Colored Petri Net)as the verification tool of UML extended model and the formal modeling specification of layered control structure,transforming the extended UML model into an executable colored petri net layered model,and through ASK-TCL algorithm checks the system model for dynamic behavior and state changes,and verifing the correctness and consistency of the UML extended model and the train controlling of RBC scenes.Finally,the possible faults in the train controlling of RBC scenes are analyzed and described according to the PHAVer(Polyhedral Hybrid Automaton Verifier)model transformed from the established UML extended model.The PHAVer and the fault model are combined to describe the train controlling of RBC scenes,the equipment failure conditions in thescenes and the impact of the failure on the function of the scenes.The fault monitor is used to calculate the reachable set of the system model and determine the combination of faults that endanger the system security so as to realize the safety analysis of the train controlling of RBC scenes.The results show that the method proposed in this thesis can meet the safety characteristics requirementsof the train controlling of RBC scenes in the demand design stage,and provide a reference method for the formal modeling and safety evaluation of the train control system.