Study On Track Circuit Fault Diagnosis System Of Railway Signal Based On Fuzzy Neural Network

The track circuit fault diagnosis system of railway signal is mainly to achieve the track circuit fault diagnosis and intelligent analysis functions,so as to provide reliable technical support to find hidden danger, ensure transport safety, fault analysis, improve the maintenance level and repair efficiency. As the main mode of track circuit in railway station,25Hz phase-sensitive track circuit is suitable for both electrified and non-electrified sections.It has many obvious advantages and its property can directly influence the transportation efficiency and safety. The slight effects of track circuit fault is to influence driving and reduce transportation efficiency,and the severe effects of track circuit fault is to cause casualties and property losses.At present,the fault diagnosis of railway signal track circuit is still not completely in the repair according to condition stage in China,and it often rely on the experience of the field staff (especially in the small station) in the process of fault diagnosis.Therefore,it is still an important project to keep on developing a scientific and intelligent fault diagnosis system to improve the security of China railway transportation and transportation efficiency.This paper combined to the scientific research project of "Research and development of China Railway Signal Equipment Fault Diagnosis Expert System" that I have took part in, took the25Hz phase-sensitive track circuit in railway station as the research object and built the fault early-warning module based on Fuzzy Neural Network due to the complex fault.Then,it entered the analysis stage and used the method based on rules for fault diagnosis.Therefore,the fault diagnosis system of25Hz phase-sensitive track circuit which based on Fuzzy Neural Network had been built.This paper mainly had done the following work:Firstly, according to the characteristics of25Hz phase-sensitive track circuit,it analyzed the work principle and the common faults and fault reasons of25Hz phase-sensitive track circuit in station based on the working conditions.It used the four-terminal network model and uniform transmission line theory to calculate the critical voltage value which would be the system’s alarm value.Finally,combined with the contents of25Hz phase-sensitive track circuit, it gave the general framework and structure design of this paper.Secondly,according to the working conditions of25Hz phase-sensitive track circuit,it put the microcomputer monitoring data input into the system for fault analysis(the analysis involved using fuzzy neural network to fault diagnosis).Finally, it outputted the faults and fault alarm according to the results of the analysis.Thirdly,in the process of track circuit analysis, the fault diagnosis model based on Fuzzy Neural Network was established for fault diagnosis.The paper let four common fault with a membership of25Hz phase-sensitive track circuit as the output of the system.Then,it used the method based on rules for fault diagnosis and output the final results(such as fault reason, fault type, fault solution and so on).Fourthly,according to the fault diagnosis system of25Hz phase-sensitive track circuit which had been built,it used C++Builder software to build the system platform,and put the microcomputer monitoring data which stored in the database into the section analysis module for the intelligent analysis and fault diagnosis.Then,the system output the results of fault diagnosis and fault alarm.