Optimal Generation Of Test Sequences For On-Board Equipment Of High-Speed Railway ATO System

With the rapid development of China’s economy and the increased demand for railway transportation,high-speed railways are increasingly becoming the primary mode of transportation for the Chinese people.High-speed railways have also gradually started to develop in the direction of intelligence and automation,and now the Automatic Train Operation(ATO)function has been successfully applied to the high-speed railway train operation control system.High-speed railway ATO system that is based on the CTCS-3(Chinese Train Control System)train control system superimposed on the function of ATO,at present by China’s independent research and development of high-speed railway ATO system has been successfully applied to the Beijing-Zhangjiakou high-speed railway,the first time to achieve a speed of 350 km per hour automatic driving,effectively improve the efficiency of train operation while reducing train energy consumption,saving resources.The functional correctness of the high-speed railway ATO system has a great impact on the safety and efficiency of traffic.Therefore,in order to ensure the correct and reliable function of the system,higher requirements are put forward for the testing work of the system.The merits of the test sequence directly affect the efficiency and level of the testing work,while the coverage and redundancy of the test cases in the test sequence determine the quality of the testing work from a certain extent.At present,the generation of test sequences is basically done manually by stringing them together,which not only leads to inefficient testing,but also may have the problem of incomplete test case coverage or too much duplicate content of test cases,and the final test results may lead to certain security risks of the system.An efficient test sequence can not only reduce the workload of manual testing,but also ensure the comprehensiveness of system testing,so the study of a high-quality test sequence generation method has a guiding role in further improving the testing work.In this thesis,based on the theory of Time Automaton,a method for automatic generation of test sequences is designed and the generated test sequences are optimized,the main research content is as follows.First,the test methods are classified and introduced based on the test sequence generation problem,and the test methods used in this thesis are clarified through comparative analysis.Based on the new content of the high-speed railway ATO system,the functional requirements of the system are analyzed,and the operation scenarios of the system are classified.Based on the theory of time automaton,the subsystems involved in the operational scenarios and the information interaction between the subsystems are analyzed,and the scenarios of linkage control of train doors/platform doors,section cross-TSRS(Temporary Speed Restriction Server)operation and automatic departure from the starting station are selected for modeling and simulation,and the BNF(Backus-Naur Form)statements are written to verify the correctness and accessibility of the system functions.Secondly,by analyzing the location information and migration information in the XML(Extensible Markup Language)document generated after modeling each operation scenario,the information obtained by parsing is stored in the cross-linked table.The test sequence generation principles are analyzed and introduced,and a test sequence generation method based on the full-variation coverage criterion is designed,through which the test sequences of the automatic departure scenarios at the origin stations are automatically generated,and the function that the generated test sequences need to cover all test cases is realized.Finally,the generated test sequences are optimized using the particle swarm algorithm to reduce the repetition rate of test cases in the test sequences.At the same time,the crossover variation idea similar to genetic algorithm is added to the particle swarm algorithm so that the particle swarm algorithm can go beyond the local optimal solution to iteratively find the global optimal solution.The redundancy of the test sequence is significantly reduced by the improved particle swarm algorithm compared with the test sequence before the optimization,and the convergence speed is effectively accelerated compared with the particle swarm algorithm before the improvement.