Power Equipment Fault Diagnosis and Prevention Based on Comprehensive Feature Quantity Evaluation


  • Bin Ma Shuohuang Railway Development Co. LTD, Suning 062350, Hebei, China
  • Xueming Jin Beijing Smart Chip Microelectronics Technology Co., LTD, Beijing 102299, Beijing, China
  • Zhe Li Shuohuang Railway Development Co. LTD, Suning 062350, Hebei, China


Electric power technology has made a vital contribution to the development of today’s society, and electrical equipment also occupies a very important position in railway operations. Since electrical equipment is very dangerous, with the improvement of living standards, people are paying increasing attention to the safety of such equipment. Frequent failure of power equipment can have serious consequences. Therefore, the important problem that needs to be addressed is how to effectively diagnose and prevent faults in railway power equipment. For this problem, the method of comprehensive feature quantity analysis can diagnose the fault of power equipment in time and effectively prevent the occurrence of the fault. Compared with the traditional fault diagnosis and prevention methods of railway power equipment, comprehensive feature quantity analysis mainly analyzed the causes of different types of faults and found out the corresponding internal components of the equipment for troubleshooting. The various types of causes were classified intelligently, and then these feature quantities were converted into data using computer algorithms to find similar variables. Lastly, the final preventive plan was obtained through model deduction. In this paper, an artificial neural network is used for data mining and analysis, and comprehensive feature quantities are examined to detect various indicators of the functioning of electrical equipment in a railway system, in order to diagnose and prevent power equipment faults. The effectiveness of the fault identification and the performance of the system were tested. It can be seen from the test results that when the load level was 80% and the number of training set samples was 68 and 65, the training set classification error rate and the test set classification error rate of the system would increase with the increase of the failure time. When the failure time was 0.1s and the load level was 80% and 100%, the response time of the system was not much different, but there was a difference of 0.82 in the mean square error value.

Keywords: comprehensive feature quantity analysis; power equipment; fault diagnosis and prevention of railway power equipment; feature mining; fault monitoring

Cite As

B. Ma, X. Jin, Z. Li, "Power Equipment Fault Diagnosis and Prevention Based on Comprehensive
Feature Quantity Evaluation", Engineering Intelligent Systems, vol. 31 no. 6, pp. 473-482, 2023.