Optimized Generation Of Test Sequences For Interface Type Test Between ATP Onboard Equipment And High-speed Train

High-speed railways in China has achieved a rapid development in recent years,which makes the safe operation of high-speed train really important.As the master control equioment,ATP(Automatic Train Protection)onboard equipment monitors the speed of train,guarantees the safety of train's operation and improves the running efficiency.There exist many interfaces between ATP onboard equipment and high-speed train.Interface type test between ATP onboard equipment and high-speed train verifies the adaptation and safety of interface,provides the original test data for the analysis of functions and performances,and enhances the type design of new standard high-speed train at the same time.It's one of the significant field tests of train control system for high-speed train.Test schedules and conducting experiment base on the test sequences of interface type test.Test sequences are now manually edited by the test experts according to test experience,types of ATP onboard equipment,test line and the test requirements.However,potential problems may exist during editing test sequence by test experts,including out of satisfaction of test requirements,the availability and the safety when executing test cases.And also,performance of manually generated test sequences cannot be qualitatively and quantitatively measured during editing process.In addition,manual work is inefficient.In order to tackle with these shortcomings,this paper studies the optimized generation of test sequences for interface type test,of which related works are summarized as follow:(1)Several basic principles,requirements and objectives are proposed for the optimized generation of test sequences according to expert experience and test requirements,including the availability and safety,redundancy and coverage,time spend and energy consumption.(2)A strategy is designed to generate optimized test sequences for interface type test,which integrates with updating process based on deep learning and optimizing process based on genetic algorithm.Models of generating test sequence are built to seek sub-optimal solution for the expected optimized test sequences depending on the comprehensive optimized objective.(3)Simulations are conducted for the designed strategy of generating test sequences based on MATLAB platform.Results show that the updating process can effectively update the improper test cases according to potential set of cases based on the pre-designed updating principles.Integrating with updating process,the optimizing process can generate optimal test sequences as expected.Comparing to the actual data of interface type test,including sequences data,train data and line data,results further demonstrate the effectiveness of proposed strategy in this paper.(4)Based on the correct parameters derived from the simulation of proposed algorithm and strategy,an application platform is programed to generate the expected test sequences by using C#programing language,which can be utilized to generate test sequences in actual interface type test.