2P15 An analysis of High-Speed Train Axle Temperature Based on Fuzzy C-Means Clustering Algorithm(Shotgun Session)

Abstract

In recent years, especially in the past ten years, with the development of high-speed railway of China, the transport efficiency has been greatly developed, and the traffic situation in China has been improved dramatically. Meanwhile, with the short-term centralized development of Chinese Train Control System (CTCS), the fault diagnosis and maintenance of CTCS will be a serious challenge in the following years. According to statistics, the mechanical failure is accounted for over 50% of the total failure, and most of them are axle failure. Axle failure may cause casualties and economic losses, and threaten the safety of high speed railway. Theoretically, axle temperature monitoring is an effective method to prevent the train accident caused by emergence of the hot axle or cut axle. Therefore, analyzing the temperature data, mastering the changing rule of axle temperature and work status of axle are of important significance. In practice, the absolute temperature of axle is usually used as a unique factor to judge the axle fault. However, due to failure to take account of the influence of dynamic factors, such as high speed, long operating distance, and environmental fluctuations, the results of fault diagnosis are poor. The feature of axle temperature is obtained through the trends curves of axle temperature and other parameter.