Research On Optimization Of CBTC Automatic Train Protection System Test Cases

The CBTC system has become the mainstream signal system for urban rail transit.The Automatic Train Protection system,as the safety core subsystem of CBTC,must be strictly tested with simulation system before going on track,because of its high reliability and availability requirements.At present,the test cases of ATP system are almost all manually compiled,with a large number and a large amount of duplicate content.In addition,the test is very tedious and time-consuming because of many meaningless and repetitive actions,when executing test cases one by one.Therefore,it is of practical significance to optimize the test cases of the CBTC Automatic Train Protection system.Firstly,the functional requirements of the Automatic Train Protection system are analyzed by this paper,according to the research of the interface and data interaction between the Automatic Train Protection system and other subsystems of CBTC.The Dream Ant Colony Optimization algorithm,DACO,is proposed to select a small number of test cases that are easy to execute on the premise of ensuring comprehensive coverage of requirements,and the reduction optimization result is compared with that of the Ant Colony Optimization algorithm,Greedy algorithm and GRE algorithm.The result shows that the number of the test case set optimized by DACO is smaller,the test cost is lower and the execution time is shorter.It proves that the introduction of dreaming factors in the update method of the Ant Colony Optimization algorithm can increase the diversity of search and improve precocious defect.Secondly,the input and output conditions of the reduced test cases are analyzed,the cascade architecture of the test sequence is designed,the key scenarios of the Automatic Train Protection system are determined,the models of key scenarios and the test driving model that conforms to the test sequence cascade architecture,are established by UPPAAL software.The test driving model can promote the conversion of the scenario models,and the test sequences are generated in the simulator.The idea of firefly algorithm being used to design the test sequence optimization process,the test sequences generated based on scenario models are optimized,and compared with genetic algorithm.The result shows that it can significantly reduce the number of reuse of test cases,and reduce the redundancy by 13.4% with the firefly algorithm optimization.Finally,the test case reduction optimization,test sequence generation and optimization method are achieved programmatically with the help of the CBTC universal test platform.The scenario library of test cases is constructed,test case reduction and optimization module and test sequence generation and optimization module are designed.The test case reduction and optimization module being used to optimize the 177 test cases for the functional testing of the Automatic Train Protection system,136 test cases are obtained after the optimization witha reduction of about 23%.The scenario model file is imported into the test sequence generation and optimization module for analysis and generation,and the two test scenario sequences,RM overspeed protection and exceeding the mobile authorization point,are tested in a simulated real site,finding that testing the optimized test sequences takes less time than testing all the test cases included in the two scenarios one by one.The results show that the test case optimization method proposed in this paper can debase test intensity,reduce test repetitive operations,improve test efficiency and save test costs while ensuring requirements coverage.