Analysis And Research On Fault Diagnosis Of Electrical Centralized Circuit Based On Decision Tree Algorithm

Due to the long history of electric interlocking,it still occupies a large proportion in the general speed railway.At present,most electrical interlocking equipment used for a long time,the average failure rate of the circuit increased a lot,although the railway sector has installed computer monitoring equipment,but the electric circuit fault diagnosis method and no substantial improvement,mostly rely on manual analysis work signal for processing the fault.Therefore,according to the electricity section electric interlocking circuit of high failure rate,manual processing time-consuming and lack of practical problems based on the experience of personnel signal,fault diagnosis technology,expert system,decision tree algorithm comparison,this paper proposes a decision tree learning algorithm on electrical fault diagnosis expert system based on the circuit,and the design of the system analysis and research and implementation.The main contents are as follows:(1)Based on the composition and structure of the electric interlocking circuit especially on network line and unit circuit principle,combined with the electricity section fault examples,the reasons of failure and treatment measures,and summed up 7 types of electric circuit fault,and the fault of 7 kinds of fine is divided into 22 kinds of phenomenon,ready for the follow-up study.(2)The basic model of fault diagnosis,the process of fault diagnosis and the structure of expert system are deeply analyzed.Built the basic model of electric circuit fault diagnosis system,according to the fault feature vector set Yi to find the match,eventually infer the fault relation set F;on the basis of the process of fault diagnosis,analysis and fault diagnosis decision fault characteristic signal and fault condition of mining.(3)In order to solve the problems of knowledge acquisition of expert system,the attributes of the decision tree algorithm selection metric analysis,C4.5 algorithm was selected as the research object,combining this method with the structure of the expert system,a fault diagnosis expert system based on decision tree algorithm system.According to the actual situation of railway signal equipment fault diagnosis,knowledge acquisition rules and knowledge base are designed,and a fault reasoning mechanism with conflict resolution strategy is designed.In this paper,the establishment process of decision tree algorithm is deeply discussed,and the construction process of decision tree for electric fault diagnosis is analyzed with examples.(4)Finally,combined with the overall structure of the fault diagnosis system of the railway signal equipment design requirements,the Client/Server system software architecture,using Java Swing as the development software,realized the electrical decision tree algorithm on Fault Diagnosis Expert System Based on circuit.The system also has typical fault case collection,knowledge base management,fault inquiry and other functions.The system is simulated and analyzed,which lays a foundation for the practical application of the system.