Research On The Measurement And Evaluation Method Of Engine Piston Barrel Shape Based On Image

The piston is a key component of the automobile engine,and the piston skirt is currently widely used with a medium-convex elliptical profile.The skirt profile of the medium-convex elliptical piston is complex and difficult to measure,and the existing measurement methods are not easy to complete the detection of the skirt profile of the medium-convex elliptical piston.In order to improve the detection efficiency of the piston skirt profile size,this paper adopts the image-based detection technology to achieve the detection of the engine piston skirt profile,and the main work is as follows:Firstly,a monocular camera-based visual inspection scheme for piston skirt contour is constructed,which divides the piston skirt contour inspection into measurement height positioning and contour edge coordinate detection.The experimental platform of the piston skirt profile measurement system consisting of CCD surface array camera,double telecentric lens,parallel light source and precision servo turntable is built according to the measurement scheme.Secondly,the sub-pixel edge detection algorithm and 3D reconstruction method of the piston skirt profile were studied.Franklin moments are used to extract the feature points of the piston skirt under test,and the height position is measured by positioning the feature points,and the grayscale value in the horizontal direction is fitted with the inverse tangent function at the specified height,and the position relationship between the real edge and the fitted curve of grayscale value is determined with the help of standard plug gauge,so as to realize the sub-pixel edge positioning.Based on the coordinate conversion relationship of the measurement system,the 3D coordinate conversion of the piston skirt edge points is completed,and the 3D point cloud model of the skirt is established.Thirdly,the profile degree error assessment method of the piston skirt was studied.According to the definition of line profile error and the principle of minimum inclusion,the least squares method is used to fit the edge profiles of the measured piston skirt in the cross and longitudinal sections,and the line profile error of the measured piston skirt in the cross and longitudinal sections is judged by calculating the maximum and minimum values of the normal distance between each measurement point and the fitted profile.According to the definition of surface profile degree error and the principle of minimum inclusion,the theoretical profile surface of the piston skirt under test is established according to the dimensional requirements of the piston design drawings,and the maximum and minimum values of each measurement point to the theoretical profile surface are calculated,so as to calculate the surface profile degree error of the piston skirt under test.Finally,the detection accuracy of the piston skirt profile of the method in this paper was verified.By comparing the measurement results of this method with those of a high-precision contact measurement instrument,it is verified that the edge detection accuracy of this method is about 1/5pixel,and the standard deviation of radius measurement is 0.0078mm.This method can realize simultaneous image processing and image acquisition,which basically ensures that all the piston skirt edge profile information can be extracted after the image acquisition is completed,realizing efficient measurement of piston skirt profile.