Research On Early Warning Method Of Urban Rail Transit Fault Based On Support Vector Machine

With the rapid development of urban rail transit in China,the related hidden trouble has become more and more prominent,which has become an important reason to affect the normal operation of urban rail transit.Urban rail transit operation companies all over the world attach great importance to the potential faults,and carry out special activities of investigation and rectification for many times.However,the hidden trouble is sudden,and the requirements for maintenance and repair of different equipment in different regions are different.If the hidden trouble is fully covered every time,the operation cost will be greatly increased;At the same time,the recording standards and classification standards for potential failures of rail transit operation companies in different regions and cities are not unified,which may lead to the situation that it is difficult to carry out standardized management and scientific analysis on the data of potential failures of urban rail transit.Therefore,it is impossible to effectively use the historical potential failures information to early warn the future potential failures,It is necessary to find a scientific method to early warn the hidden trouble of urban rail transit,so as to accurately and timely guide the maintenance.Aiming at the problem of early warning of urban rail transit hidden trouble,this paper puts forward a classification method according to the existing hidden trouble,and combines the hidden trouble data obtained by crawler technology with the official data provided by urban rail transit operation company to extract the effective information related to the hidden trouble,Support vector machine(SVM)model and BP neural network(BP neural network)model are used to carry out the early warning experiment of urban rail transit fault hidden danger,and a case is formed to guide the maintenance of the operation company.By effectively grasping the occurrence mechanism of the hidden trouble,this paper puts forward the fault information collection and response methods suitable for the operation of urban rail transit to solve the hidden trouble,and provides guidance for the operation and maintenance of urban rail transit.The main tasks are as follows:(1)According to the current national standards,the current regulations of many urban rail transit operation companies and the results of field investigation,with the help of the urban rail transit fault hidden danger data obtained by crawler technology and the official data provided by urban rail transit operation companies,a classification and classification method of fault hidden danger is proposed,which provides the basis for the future work of fault hidden danger early warning.(2)Based on the processed data of urban rail transit fault hidden danger,the early warning experiment of fault hidden danger is realized based on the support vector machine model.In the case of different dimensions and different amount of data,the effect of model operation is analyzed to find out its impact on the accuracy of the model,so as to determine the best data dimension and the best amount of data;On this basis,the paper tests the early warning effect of support vector machine model and BP neural network model,and finds that support vector machine model is more suitable for urban rail transit fault early warning.At the same time,according to the further needs of early warning work for data,this paper puts forward the information collection method of urban rail transit fault hidden danger,points out the problems existing in the troubleshooting and governance of the existing problems,puts forward the corresponding solutions for the existing problems,and designs the troubleshooting and Governance methods for different types of faults.It can not only collect data more accurately and in detail,but also implement the work responsibility of fault handling.It can not only solve the current faults,but also apply the collected data to the early warning of future faults,so as to better ensure the normal operation of urban rail transit.(3)According to the early warning results and troubleshooting methods,combined with the state perception analysis of urban rail transit,the connection between the early warning of hidden trouble and intelligent operation and maintenance is explored.Through the case application,the combination of the early warning results and the state perception of urban rail transit system can play a certain guiding role in the future development of intelligent operation and maintenance of urban rail transit.