Fault Diagnosis Of ZPW-2000A Track Circuit Based On Decision Tree Algorithm

The safety of railway transportation needs many aspects of coordination and cooperation, and the quality of equipment is one of the important aspects. As the main railway technical equipment, the railway signal equipment play a key role in ensuring traffic safety and improving the efficiency of railway transportation. Therefore, in order to ensure the safety of railway transport, one of the key points is to ensure the normal operation of railway signal equipment.At present, the telecommunication department use the centralized signaling monitoring (CSM) to monitor the railway signal equipment in overall. However, the railway signal equipment is numerous, the structure and logic is complex, and fault analysis and maintenance still rely on manual. In order to improve the efficiency of railway signal equipment fault diagnosis, it is very necessary to develop a set of fault diagnosis system to solve this problem.Based on the above purpose, the paper selects ZPW-2000A jointless track circuit as the research object, which is widespread used in the railway signal field. Firstly, the thesis analyses the structure and principle of ZPW-2000A jointless track circuit, and it also analyses the principle of real-time monitoring of ZPW-2000 A jointless track circuit data collected by CMS system, types of acquisition and monitoring data, and alarm information. Combined with the general flow of fault diagnosis, the fault diagnosis process of ZPW-2000A jointless track circuit is summarized.Secondly, the main machine learning methods are analyzed and compared. According to the characteristics of ZPW-2000A jointless track circuit, the paper use the C4.5 decision tree algorithm as the method of knowledge acquisition, and optimizes the process of discretization of continuous attributes to improve the operation efficiency. Based on the structure of the expert system, the decision tree and expert system are combined to form a fault diagnosis system based on the decision tree algorithm. Then the fault characteristic parameters of ZPW-2000A jointless track circuit are analyzed and summarized, and the overall structure of the fault diagnosis system is designed. Besides, reasoning machine, knowledge base, the knowledge acquisition, and other modules of the fault diagnosis system are analyzed and designed exhaustively.Finally, ZPW-2000A jointless track circuit fault diagnosis system is developed under the Python development environment and by combining the C4.5 decision tree algorithm, expert system and ZPW-2000A jointless track circuit. According to the structure of the fault diagnosis system, the function and flow of each module are analyzed. When the relevant data are input to the diagnosis system, the diagnosis system will give the diagnosis result and the corresponding fault handling advice, which help signal maintenance personnel to solve the fault of the ZPW-2000A jointless track circuit more quickly and ensure the safety of railway traffic.