Research On Vertical Shaft Detection System Based On CMOS-Camera And Lidar

As the throat passage of coal mining,shaft plays a pivotal role in coal mine safety production.Before and after production,shaft wall deformation and fracture due to axial compression and stretching,center line bending and movement,radial compression,and other reasons have become a serious mine disaster.Therefore,daily inspection of shaft is an important means to ensure coal mine safety production.In view of the existing shaft detection methods,which sometimes detect abnormalities after shaft damage,there are common problems such as complex testing systems,low intelligence,and poor long-term stability.Therefore,this article establishes a shaft detection system based on three indicators: shaft deformation,shaft perpendicularity,and horizontal vibration of the lifting container,thereby improving the automation,intelligence,and efficient detection level of the shaft,and reducing the probability of significant damage to the shaft,It has field application value and theoretical research significance to ensure the safe operation of the shaft.Firstly,based on the actual working conditions and purposes of the shaft detection system,the key functions such as data acquisition,data transmission,and data processing that the system should have during shaft detection are determined.The overall design of the system scheme is carried out.In terms of hardware,lidar,CMOS camera,laser collimator,main control board,and battery are selected as the main hardware of the system.In terms of software,a topic communication model for publishing subscribers based on ROS is determined according to the hardware characteristics,and the software implementation of communication nodes,camera acquisition nodes,and data processing nodes is completed.Secondly,an experimental platform simulating the mine shaft was built for experimental data collection.For horizontal vibration displacement detection of lifting containers,first,the camera calibration parameters were determined using the Zhang Zhengyou chessboard method,and the calibrated COMS camera was used to collect the original spot image.The image processing processes of graying,median filtering,binarization,canny edge extraction,and centroid method were determined to ultimately obtain accurate spot center coordinates.Within the vibration displacement range of the lifting container,experiments are conducted to verify the accuracy of the center coordinates of the laser spot projected to different positions in the camera’s field of view,determine the static displacement error of the visual system,and obtain the conversion parameters from pixel coordinates to real-world physical world coordinates.Thirdly,for shaft deformation and perpendicularity detection,using the established shaft detection system,under static conditions on an experimental platform simulating the shaft,collect shaft point cloud data from different angles,determine the processing flow of voxel grid down sampling,radius filtering,4PCS algorithm coarse registration,ICP algorithm fine registration point cloud data,and finally obtain shaft deformation detection and perpendicularity detection data.Collate and compare the centroid coordinate errors before and after registration,Verify the registration accuracy and determine the verticality detection error.Finally,in order to analyze the detection accuracy of the shaft detection system under actual working conditions,a speed generating device was built to provide a known constant linear velocity and then convert it into the vibration displacement of the lifting container.Experimental parameters were set to start data collection,and the detection accuracy performance of the system under actual working conditions was analyzed,verifying the reliability and practicality of the system.There are 78 figures,15 tables and 91 references in this thesis.