Font Size: a A A

Research On 3D Laser Measurement Technology For Typical Features Of Aviation Components

Posted on:2020-08-03Degree:MasterType:Thesis
Country:ChinaCandidate:X LiFull Text:PDF
GTID:2370330590993846Subject:Engineering
Abstract/Summary:PDF Full Text Request
Aviation components inevitably produce certain dimensional errors in the manufacturing process,and the accumulation of errors in different links will have a major impact on the assembly performance of aviation equipment.However,for a long time,the detection of parts mainly relies on manual work,and there are problems of poor consistency of detection quality and low detection efficiency.At present,the manufacture of aviation components is moving towards digitalization,and the digital measurement of part features is an important part.Based on the 3D laser measurement technology,the key technologies of multi-line laser measurement system model construction,multi-line laser binocular matching based on homography matrix,skin processing boundary measurement based on parallel multi-line laser,and hole-based feature measurement based on cross laser are deeply studied in this paper,and it realizes the digital measurement of the typical characteristics of aviation components.The main research contents are as follows:(1)In order to measure the machining boundary,armpit and hole characteristics of aviation skin parts,it established a multi-line laser measurement system model.Aiming at the problem that the precise target need to be produced and the extracted light plane feature points are few in the system calibration process,it studies and implements the multi-line laser plane induced homography matrix calibration method.The experimental results show that the laser plane accuracy reconstructed by homography matching can reach 0.0048 mm,and the calibration accuracy is high.(2)In order to solve the ambiguity problem of multi-line laser binocular matching,a multi-line laser binocular matching method based on homography matrix is proposed.For the parallel multi-line laser,combined with the method of laser line segment coincidence,the mismatching problem of the corresponding points of the image is further eliminated.The experimental results show that the precision of the point cloud of the stepped block with multi-line laser matching reconstruction can reach 0.065 mm,and the reconstruction accuracy is high.(3)Aiming at the boundary feature of aviation skinning parts,parallel multi-line laser is used for measurement,and the template matching method is proposed to extract the point cloud of the skin box.Firstly,the point cloud model is decomposed by custom template,then the boundary feature points are extracted uniformly for the section line obtained by decomposition.Finally,the discrete feature points are interpolated and the boundary feature lines are automatically generated.The experimental results show that the measurement accuracy of the skin processing boundary can reach 0.095 mm,and the measurement accuracy is high.(4)For the characteristics of the armpit and the circular hole,the cross-laser is used for measurement,and the geometric segmentation method is proposed to extract the feature of the crossed laser point cloud.By reducing the dimension of the point cloud model,the features are detected in a two-dimensional space,thereby transforming the three-dimensional problem into a two-dimensional problem.The experimental results show that the measurement accuracy of the diameter of the hole can reach 0.069 mm,the measurement accuracy of the axillary depth can reach 0.068 mm,and the measurement accuracy is high.(5)The software and hardware system platform is built,and the multi-line laser 3D reconstruction and part feature extraction methods studied by the standard sample are selected.The experimental results show that the measurement accuracy of the skin processing boundary and the circular hole basically meets the measurement requirements,and the measurement accuracy of the axillary depth also reaches the micrometer level.
Keywords/Search Tags:Part feature, laser measurement, Multi-line laser matching, Point cloud model, Feature extraction
PDF Full Text Request
Related items