智能故障诊断技术在地铁车辆维保中的运用与探讨

打开文本图片集

Application and Discussion of Intelligent Fault Diagnosis Technology in Metro Vehicle Maintenance

SU Xianhui1，HANG Yongxian² （1.HefeiRail TransitGroupCo.，Ltd.，Hefei23ooo1，China; 2.CRRC Nanjing Puzhen Rolling Stock Co.，Ltd.，Nanjing 21oo31，China）

Abstract：In the process of urban rail transit vehicle maintenance，traditional metro vehicle fault detection mainly relies on manual inspection，which has the problems of low eficiency and strong subjectivity. In view of this situation，this paper puts forward an innovative solution，that is，byal-round collection and analysis of existing vehicle information，establishing a standardized template identification database，comparing the newly collected vehicle information data with the identification database，andusing big data and wavelet processing technology to identify train fault information diagnosis，so as to realize theuseof machine detection instead of manual detection in dailydetection of metro vehicles. This paper introduces the processing flow，principle and keytechnologyof inteligent fault diagnosis technology in detail，and analyzes the technical characteristics and difficulties of intellgent fault diagnosis technology basedoninformation analysisand comparison in practical application，and discusses its application in metro vehicle maintenance. Finally，taking Guangzhou Metro Line 2l as an example，the application efect of this technology in actual metro maintenance is demonstrated.The system can cover 78% of the detection items，and the detection accuracy rate is as high as 98.87% . Compared with traditional manual detection，inteligent fault diagnosis technologycan significantly improve the detection accuracy，playan important role in train safety protection，and help to realizea standardized，informationbased and automated metro vehicle maintenance operation mode.

Key words：metro vehicle；inteligent fault diagnosis；image recognition； object segmentation；big dataanalytic； template identification database;wavelet processing

1车辆维保现状分析

目前，国内地铁列车故障检测和预防大多是采用人工巡检的方式，处于定期维修与故障维修的模式。（剩余4304字）