Research On Automatic Generation Method Of Test Sequences For CBTC Zone Controller Based On Model

With the rapid development of radio technology,the Communication-Based Train Control(CBTC)system based on precise positioning technology and high-speed,real-time communication technology has become the development direction of rail transit operation control systems.Zone Controller(ZC),as the core equipment of the CBTC system,must be equipped with extremely high security and reliability,so strict functional and performance tests must be conducted before being put into use.At present,test sequences used in testing ZC are still generated by manual methods.Due to the uncertainty of human factors and the fact that it is difficult to test the completeness of the system by manual test sequences,the reliability and security of the system can not be guaranteed thoroughly.Therefore,it is of great significance to study the automatic generation method of ZC test sequences.Based on the formal model of ZC,the automatic generation method of test sequences is studied in this thesis.Firstly,the function of ZC is analyzed,and the timed automata model is established.The formal modeling creates conditions for the automatic generation of test sequences because it eliminates the ambiguity of natural language and facilitates computer recognition.Then UPPAAL is used to simulate the timed automata model,and according to the functional and real-time requirements of ZC,the BNF verification statement is written to verify the model,which provides the possibility for generating correct test sequences.Then,the format characteristics of the timed automata model file are analyzed.According to its specific information storage format,the key words are extracted to obtain the state information and transition relation of ZC,and test cases are generated by combining them.Then,based on the functional features of ZC,the generated test cases are serially connected into a test sequence by using priority settings and repetitive flags that conforms to ZC workflow and test rules.Finally,the application program is written based on the rich function library resources of Microsoft Visual Studio.The application program realizes the automatic extraction of the model information after importing the timed automata model of ZC,and the automatic generation of test sequences after input initial state and end state.By comparing the automatically extracted model information of ZC with the information extracted manually,the results show that the information extraction method designed in this thesis can complete the full coverage of the model transition,which can prove the completeness of test cases.And automatically generated test sequences are consistent with the result generated by the manual connection.The results show that the generated test sequences can reduce the repetition rate of test cases to the minimum while ensuring full coverage.And the automatic generation of test sequences can effectively avoid the uncertainty risk in the process of artificial generation.