Method Of Data Safety Processing And Verification Of CBTC System For Urban Rail Transit

As a kind of automatic train control system used for safety assurance, CBTC systems has advantages of computer-based control, moving block tracking and closed loop control, etc. Compared with the traditional fixed block train automatic control system, CBTC systems describe the real system environment and drive the system function with different kinds of data. Data safety is an important aspect of CBTC system, but how to ensure data safety is a weak link of rail transit. The research of data safety in CBTC systems is of great significance. Based on the independently developed CBTC system of China, this thesis focuses on how to ensure data safety in CBTC systems. The author establishes the data-driven model and summarizes the characters of CBTC system, which includes large number of CBTC system data objects, safety constraint relationship among the data and data aging problem, etc. According to these characters, we propose a method to ensure the correctness of the restricted safety relationships among data to ensure the data safety. Then, on the basis of analysis on data error type, we study the data processing and data verification. The main achievements and innovations are shown as follows,(1) According to the large number of static data objects, this thesis proposes a method to describe structured data. Among static data for topology rules, attribute rules proposed by SAT methods for static data are processed to solve the problem of the static data errors and false data values. At the same time, when VOBC subsystem and ZC subsystem calculate the position and MA respectively, this thesis present topology description method for the system and constraint conditions of path search algorithm, to solve the problem of topological logic errors.(2) In this thesis, the operating train model is analyzed to extract the train position and mobile authorized data to study the treatment of safe location. At the same time, on the basis of the analysis on the model, this thesis introduces the performance and constraints of functional safety mobile authorization and deals with safety constraints by the LTL method.(3) For the relationships among the static data, this thesis proposes a data safety verification method which is based on CSP model. The proposed model is based on UML source model and it can transform time-series UML model into CSP model through the LTS model semantic conversion and transformation rules. Besides, an example is given to illustrate the feasibility of the model transformation method.(4) For the dynamic data dependencies that exist in the aging characteristics, this thesis presents a data safety verification method which is based on TA models. According to model transformation rules, this thesis transforms UML timing model into TA model. It solves the consistency problems of the dynamic data and the application of composite error problem when using of the data. One example is given to illustrate the feasibility of the model transformation method.(5) To ensure the safety of train, this thesis chooses platform of Tongjinan station and part of range line to do static and dynamic data safety validation test. This thesis uses aging data and safety constraints to verify the static data and uses data consistency, data aging and data safety constraints to verify the dynamic data. The result illustrates the feasibility of formal model-based approach on data safety verification.