Consistency And Interconnection Analysis Method Of ATO System In High-speed Railway

With the rapid development of the high-speed railway,the improvement of the automation level of the high-speed railway has become the focus of the current research.In such a case,the ATO(Automatic Train Operation)system arises at the historic moment.The opening of the intelligent high-speed railway marks China's high-speed railway ATO system up to higher level.All high-speed railway ATO system manufacturers do the research according to specifications issued by China Railway Group.Ideally,the equipment is completely consistent with the specifications and has good connectivity.However,due to different understanding of the specification by different developers,there will be inconsistencies between the equipment and the specification or the situation that the equipment cannot be interconnected.It is necessary to verify the consistency of equipment and specification and the interconnection between different manufacturers to ensure the system safe and reliable in the laboratory test.The traditional consistency and connectivity analysis methods mainly rely on expert experience.It is necessary to propose a reliable,efficient and effective analysis method to improve the level of automation and standardization.In this paper,a consistency analysis method based on formal modeling and system level fault diagnosis and an interoperability analysis method based on machine learning algorithm are proposed.First of all,a method of modeling and verifying operation scenarios combining UML(Unified Modeling Language)with CSP(Communication Sequential Process)is proposed.Combined with the relevant specifications of high-speed railway ATO system,this method uses UML use case diagram,class diagram,object diagram and activity diagram to describe the scenes of high-speed railway ATO system,and uses CSP to describe and verify the model according to the transformation rules.Taking the train cross TSRS operation scenario as an example,the interaction rules between concurrent system processes can be obtained.Then,a consistency analysis method based on system level fault diagnosis theory is proposed.Based on the theory of node mutual test,there will be the mechanism of node mutual test together with extracted system nodes and the formal model of operation scenario.The system level fault diagnosis model is built.After computing the test result matrix,the abnormal nodes can be located accurately using the diagnosis algorithm.Taking the cross TSRS operation scenario as an example,the consistency analysis comes to a result.Finally,SVM(Support Vector Machine),a machine learning domain classification algorithm,is used to analyze the interoperability of ATO system of high-speed railway.And the platform door linkage scenario is taken as an example to demonstrate.The main steps include the acquisition of original data,data preprocessing,SVM classifier training and verification,and the conclusion of interoperability.The least intra class variance method is used for feature selection in the process of data preprocessing.The super parameters are optimized by Bayes to improve the classification accuracy.Eventually,the performance of SVM classifier is evaluated.Given SVM classification results,we can draw a conclusion about the interoperability of platform door linkage scenario in the highspeed railway ATO system.Figure 45,Table 26,63 References.