Health Status Detection Of Pantograph Slides In Rail Vehicles Based On Image Analysis

During the operation of the electric locomotive,the pantograph slides in contact with the catenary to receive current.The quality of the slide will directly affect the stability of the electric locomotive.Therefore,in order to reduce the possibility of pantograph accidents and improve the stability of train operation,which are very important to detect the health status of pantograph slides in high frequency.This paper takes the pantograph slides of rail vehicles as the research object,and uses image analysis technology to detect the characteristics of pantograph slide wear and crack health.Firstly,based on the understanding of the working principle of pantograph,the maintenance project of slide plate,SJ pantograph net and roof state detection system,this paper introduces the scheme of slide plate detection system of pantograph,completes the selection of hardware part of the system.The conversion coefficients of pixel coordinates and spatial coordinates are obtained,through calibration experiment on selected cameras analyzed by Matlab software.Secondly,the original image of the skateboard was filtered and enhanced to obtain a clearer skateboard image.The improved Canny algorithm was used to extract the edges of the skateboard wear edge images collected by the three wear cameras,and the three wear curves were obtained through curve fusion technology.The fusion was performed to obtain the health status characteristics of pantograph slide wear.Then binary processing and corrosion operation are carried out on the pre-processing crack image to get the crack skeleton.Based on the support vector machine,the crack skeleton and the skateboard dirt are classified and trained.Then,the skateboard crack skeleton is correctly identified and the crack length is calculated.Finally,based on the Visual Studio software,the pantograph slide health detection system is designed to complete the wear and crack detection of the pantograph slide.Combined with the skateboard images collected on site in normal weather and rain and snow weather,the artificial detection results are compared.The standard value of the experiment is to verify the accuracy of the system’s slide wear detection accuracy and crack recognition accuracy.The experimental results show that under normal weather and rainy and snowy weather,the slide wear detection accuracy is 94% and 90% respectively,and the system detection errors are within ± 0.3mm,the accuracy of crack recognition in normal exposure environment and overexposure environment is 93.3% and 80%,respectively,and the crack length detection error is within ± 0.3mm.In this paper,the detection accuracy of the system is higher than that of the field equipment,so it can meet the detection requirements of the slide plate crack and abrasion.