| In recent years,with the continuous progress and rapid development of high-speed railways,railway safety has gradually attracted people’s attention.In the important part of the electrified railway,the working condition of the contact network can directly affect the safety of railway operation.Among them,the contact line is in direct contact with the pantograph’s carbon slide,and the contact force between the two is used to maintain the current supply of the contact line to the pantograph at any time to ensure safety during driving.Therefore,the frequent detection of the contact pressure between the pantograph and the catenary is of great significance to the normal operation of the railway locomotive.Before the train leaves or when the trip reaches a certain distance,the condition of the pantograph and the roof of the train must be detected.The static contact force of the bow net is one of them.At present,most of the detection methods adopted by most railway bureaus are contact.The staff stands on the roof and pulls down the pantograph with a spring dynamometer.The reading displayed on the dynamometer is the current contact force.This detection method neither guarantees the safety of the staff,but manual reading of the data will cause reading errors with low accuracy.In view of this situation,this paper will adopt a non-contact detection method and design a set of pantograph static contact force detection system based on monocular vision.The core content of this system is to use image processing technology to lift the contact line according to the pantograph.The contact line displacement at this time detects the current static contact force of the pantograph.First,the overall structure of the system is introduced,including the data acquisition subsystem,image acquisition subsystem,and data processing system,and the hardware equipment and installation methods of each subsystem are introduced in detail.Among them,the data acquisition subsystem uses sensors to collect static force and contact line displacement data,and is transmitted to the PC through the signal processing circuit and single-chip analog-to-digital conversion.The Gauss-Newton-polynomial method is used to fit the characteristic relationship between the two.The image acquisition subsystem The CCD camera is used to collect the image of the change of the contact line,and the displacement of the contact line is determined by the pixels.Secondly,the camera calibration is carried out using Zhang’s calibration method.According to the calibration experiments,parameters such as internal and external parameter matrices and distortion coefficients of the camera are obtained.The camera parameters are used to correct the image.Then,according to the actual application scenario and measurement technical requirements,a measurement scheme for contact line displacement is proposed.The method of combining feature point matching based on ORB and feature line matching based on LSD is adopted,and feature matching is performed using KNN method.In order to solve the problem of mismatching in matching,PROSAC method is used to remove mismatching,and then the line coordinates after matching are completed.The coordinate range of the point is constrained to calculate the coordinates of different positions of the contact line.The actual distance of the contact line displacement is calculated based on the camera calibration results.The static force is calculated according to the characteristic curve of the static force and the contact line displacement.Finally,the measurement accuracy of the entire system was verified,and the system’s existing errors were analyzed and analyzed.Pycharm was used as the development platform,and the Open CV library was used to complete the system programming in Python.Pyqt5 was used to develop the interface of the entire system designed to realize data display,calculation,storage,query and other functions. |