Research On The Hardware Reliability Analysis Method Of The High-speed Railway Centralized Traffic Control System

The centralized traffic control system,as the core technology and equipment of railway transportation,plays an important role in ensuring the safety,high speed and punctual operation of trains.Hardware equipment is an important infrastructure of the centralized traffic control system,its reliability directly determines the system’s operation safety,reliability and stationary.Reliability analysis of hardware and analysis of hardware failure mode and failure mechanism,are significant to improve the reliability of the centralized traffic control system and ensure the quality of high-speed passenger transport service.The hardware composition of centralized traffic control system in our country is complex.There are many kinds and large quantities.It is very difficult to model the system through traditional reliability analysis method for extremely difficult calculation and inadequate data application.The development of railway digitization accumulated large amount of on-site reliable data,that is,on-site reliable data,which provided data support to the application of data mining algorithm to analyze hardware reliability.The research contents are the following:Firstly,the architecture and hardware equipment of centralized traffic control system were introduced,and the characteristics of the complex hardware structure,the large variety and quantity of the equipment are depicted.The implementation strategy of data mining technology in centralized traffic control system hardware reliability analysis was described based on its hardware structure.In addition,the function and architecture of hardware reliability analysis module of centralized traffic control system were designed.The basic flow of reliability analysis methods based on data mining is expounded according to the combination of the basic steps of data mining with the basic flow of existing reliability data analysis methods.Secondly,this thesis took the hardware equipment of the station autonomous computer as an example,analyzed the characteristics of hardware failure and its failure factors,also the uncertain relationship between them;analyzed the method of collecting the on-site reliability data of the hardware of station autonomous computer based on the architecture of centralized traffic control system and hardware composition;designed elaborated the logic framework of on-site reliability data preprocessing based on the application framework of railway big data.After compared the performance of Beijing-Zhangjiakou high-speed railway and Beijing-Xiongan high-speed railway from the same type and batch of hardware equipment of autonomous computer,whose failure factors including design level,manufacturing process and running state,installation and debugging personnel are under similar conditions,we found that there is significant difference in running environment factors.Therefore,autonomous computer’s reliability data set based on environmental failure factors was constructed to provide data support for hardware failure factor analysis of autonomous computer based on data mining.Lastly,according to the identical notion of similarity relation and similarity degree of clustering algorithm and rough set theory based on indistinguishable relation,failure factor analysis and decision model of hardware failure of station autonomous computer was constructed,the failure decision model of CPU boards of the station autonomous computer was constructed,whose condition attribute set is the environment of machine room.And equivalence class was divided into the hardware field reliability data set of the autonomous computer based on environmental failure factors.From the perspective of single environmental factor and coupled environmental factor,the approximate degree of partial dependence between environmental failure factors and hardware failures of the autonomous computer was verified and analyzed,and the construction suggestions for improving the hardware reliability of the autonomous computer were provided.The feasibility and applicability of the hardware reliability analysis method based on data mining are verified through completing all the basic processes of the reliability analysis method.