Research On Galloping Monitoring Of Catenary Positive Feeder Of High-speed Railway Catenary In Gale Area Based On Image Processing

The Lanzhou-Urumqi High-speed Railway passes through the five major wind regions of Gansu and Xinjiang.It has a long traffic mileage and a complex environment.The catenary positive feeder galloping frequently along the line.The galloping causes frequent disconnection,interphase discharge,and hardware wear and other accidents.In the traditional mode,the galloping monitoring mainly relies on manual inspection,and the uncertainty factor is large.Therefore,by designing a catenary positive feeder galloping monitoring system,accurately identifying the catenary positive feeder target in the video image,and analyzing the amplitude,frequency and other data of the catenary positive feeder galloping,it is possible to give early warning to the status of the railway along the line,and improve the ability of intelligent monitoring and management of the railway.Firstly,this thesis designs an on-line monitoring system for galloping catenary positive feeder based on video image processing based on the distribution characteristics and online monitoring requirements of the catenary positive feeder of Lanzhou-Urumqi High-speed Railway catenary,and selects the appropriate location for the camera arrangement.Through the classification and statistics of a large number of catenary positive feeder images,the characteristics and distribution of catenary positive feeder in the images are classified.Considering the complex environment and background of the catenary positive feeder,the surrounding pixels of catenary positive feeder in the image under different backgrounds are counted,and the threshold range of offline feature detection in different environments is obtained from the three-dimensional pixel distribution of the catenary positive feeder and the pixel values obtained by statistics.Secondly,preprocess the image of the catenary positive feeder,and combine the characteristics of the catenary positive feeder in the image to count the grayscale stretch intervals of the catenary positive feeder in different types of images,compare the effects of various preprocessing algorithms and the pros and cons of the steps,and design a reasonable and fast image preprocessing step to complete the enhancement processing for the catenary positive feeder target in the image.Considering the difference in the environment where the catenary positive feeder is located,an appropriate edge detection algorithm is selected,the obtained line feature detection threshold is introduced,and the Ratio operator is used to process the preprocessed and enhanced image,and at the same time,the catenary positive feeder edge and the background noise classification method completes edge extraction and verifies the accuracy of the algorithm.Finally,extract and analyze the galloping monitoring data.According to the principle of visual calibration,relying on the principle of galloping monitoring,the coordinates under different coordinate systems are transformed,an experimental physical simulation platform is built,the internal and external parameters of camera calibration are obtained,and the target world coordinates are obtained.Calculate the relative amplitude of the movement of the feature points in the image,analyze the vertical and outward movement of the target measuring point,and complete the frequency calculation according to the camera shooting law and the law of the characteristic direction of the galloping.The Lanzhou-Urumqi High-speed Railway catenary positive feeder galloping state for warning.