| As a rail vehicle,the bogie is the most important part of the railway freight train,and the measurement of the bolster spring size and wedge wear of the bogie is the most important link in the bogie inspection.Whether the bolster spring and wedge meet the requirements of the inspection regulations will directly affect the stability of the vehicle operation.In the process of traditional bolster spring size measurement and wedge wear measurement,it mainly depends on manual use of corresponding non-standard measuring tools.In order to solve the current situation that the measurement of bolster spring and wedge depends on manual work,this paper studies and designs the bolster spring wedge measurement system for K2/K6 bogie based on machine vision technology and the existing maintenance line.In this paper,the overall design scheme of the detection system is described in detail,and the corresponding measurement algorithms are designed according to the specific measurement requirements of bolster spring size and wedge wear.Firstly,the measurement requirements and actual working conditions of bolster spring and wedge are analyzed,the task requirements and design indicators of bolster spring wedge measurement system are clarified,and the bolster spring wedge measurement system with optical mechanical electrical integration is studied and designed.In the research and design process of this system,the key line laser sensor is selected and verified;The system reliability,maintainability and safety shall be fully considered and corresponding design measures shall be taken.Further,the detailed design and type selection of the system structure are further completed according to the design indexes determined by the analysis.In the process of system structure design,the key data acquisition module is designed and checked;Calculate and select each module;The key structure of the system is simulated and analyzed.Through calculation and verification,the selection and design of the system meet the design requirements.Secondly,according to the design indexes determined by the analysis,the detailed design and selection of the system structure are further completed.In the process of system structure design,the key data acquisition module is designed and checked;Each module is calculated and selected;The motion and structure of the components are verified in the simulation.Finally,the measurement algorithm is tested.In the test,the repeatability accuracy of the measurement results is ±1.5 mm and the measurement time is 60 s.Thirdly,the bolster spring size measurement algorithm based on visual image is studied and designed.Through the line laser sensor in the image acquisition system for data acquisition,the gray image containing height information is obtained.According to the distribution law of gray value in the image,the free height of bolster spring is calculated.Furthermore,based on the gray level straight square threshold segmentation method,the shadow contour interference is filtered out,and the accurate edge contour of the bolster spring is obtained to measure the inner diameter and outer diameter of the bolster spring.Fourthly,based on the analysis of the requirements of wedge wear measurement regulations and the characteristics of measurement environment,a wedge wear measurement algorithm based on point cloud is studied and designed.The measurement algorithm includes point cloud preprocessing algorithm,outlier removal,straight through filtering,plane fitting and wedge wear measurement algorithm.Combined with the experiment,the optimal parameters of the measurement algorithm are determined,and the automatic measurement of wedge wear is realized.The repeatability accuracy of the measurement results is ±0.35 mm and the measurement time is 80 s.Finally,according to the designed bolster spring size and wedge wear measurement algorithm,based on Windows platform,taking Visual Studio 2019,Qt5,Open CV4 image algorithm library and PCL11 point cloud algorithm library as the development environment,the software design of bolster spring wedge measurement system is completed.The measurement software has a simple human-computer interaction interface. |