The Key Techniques Research And Application On Automation Test For Train Control System In Urban Rail Transit

Urban rail transit has developed rapidly and is becoming the preferred tranvel way for people because of its characteristics of large transportation capacity,energy saving,environmental protection,convenience and safety.As its control center,the train control system in urban rail transit is the key of ensuring the safe travel of passengers and the efficient running of trains.The large transportation capacity,the high-density driving and extremely short intervals have brought challenges to the train control system,and it is necessary to carry out long-term,fully covered system tests and field engineering tests.Communication-Based Tain Control(CBTC)is the most advanced system in the train control system for urban rail transit,which is a highly complex safety-critical system and is highly dependent on the unique rail line environment.Affected by the urban rail transit construction period,it is difficult to carry on the long-term field test.Therefore,it is necessary to research the automatic testing technology for urban rail traffic train control system to reduce the field testing time.This thesis relies on the national Science and technology support plan,combined with the engineering projects of the company working for a Ph D in engineering.It summarizes the shortcomings of the existing urban rail transit train control system automatic testing.And what’s more,A CBTC system automatic testing scheme is proposed which is based on the combination of virtual and real combination.Firstly,starting from the accurate simulation of the CBTC system’s control object,a method of constructing a multi-condition train simulation model driven by real data is proposed,so as to replace the real train with the accurate train simulation model as the control object of the CBTC signal system.Then,the timing and positon coupling control characteristics of the CBTC system are analyzed in depth.So,the reduced and efficient complex scenario test cases for the CBTC system are automatically generated.Then,the hierarchical semantic drive technique is adopted to generate the test case scripts according to scenario test case matrix,combined with the test sequence generated by the shortest time cost algorithm,the test suit is quickly executed.Finally,an automation testing platform integrated the above key techniques for the train control system in urban rail transit is developed and realizes engineering application.The innovations of this thesis are as follows,(1)In view of the inability to combine the train running simulation model with the train actual driving in test site,and it is difficult to take into account the real-time performance and accuracy of the train driving simulation,a method for constructing a multi-condition train simulation model driven by real data in spot test is proposed,and the multi-condition train running simulation model based on the real train driving characteristics is established.On this basis,a train parameter identification model based on the test field train driving data is established.By using multiple group genetic algorithm to optimize the interval of variable forgetting factor of AM-VFF-RLS based train time-varying model,it realizes the accurate simulation of the real train operating conditions in field engineering tests.(2)For the heavy workload of handwitten for the scenario test cases of the CBTC system,and the difficulties to select the time of scenario test,and it is hard to perform the multi-scenario concurrent test,the CBTC system test case generation technology based on timing and position coupling control is proposed.At First,a train control system model based on timing and position coupling is established.Then,the frequency and scope of timing and position coupling control of CBTC system is used as the evaluation to analyze its key control characteristics,and the key control feature points are extracted.Then,an automatic test model of the operation scene based on the control feature extraction and random feature fusion is generated.Finally,the method based on the safety characteristics of the CBTC system is used to reduce the generated multiscenario concurrent test cases,and the treamlined and effective scenario test case matrix is provided for testing platform for the train control system in urban rail transit.(3)For CBTC system scenario tests largely rely on manual test cases sorting and interface operation,it is difficult to carry on automation test,the automatic execution technology of train control system test based on hierarchical semantic drive is proposed.At first a hierarchical semantic model is constructed according to the complex interface transmission characteristics of the CBTC system.Then,the keywords are hierarchically designed and generat the scenario test case scripts on the basis of semantic model.the hierarchical semantic model is implemented,and the test case scripts are generated.Then an automated testing framework for the general application of CBTC system is built.Finally,the test sequence generation algorithm based on the shortest test time cost is used to connect the test case scripts in series,run them automatically in the test environment of virtual and real combination,and output the test report.(4)For the engineering challenge that the safety-critical CBTC signal system relies on on-site environmental testing and is difficult to achieve rapid delivery,an automated testing platform based on the combination of virtual and real combination for the CBTC system is developed,which was used to test Hangzhou Line 6,Ningbo Line 5 and other subway lines.The hybrid test of 11 subway lines in 2020 and 2021 has been carried out,which greatly accelerates the delivery of the CBTC system.The delivery cycle is shortened by a third,which verifies the versatility,scalability,reliability and speed of the automatic test system for CBTC system developed in this thesis.