Research On Surface Information Generation Method Of Surface Parts Reverse Engineering

With the development of industrial technology,curved parts are widely used in industrial machinery and equipment.The rapid and accurate generation of surface information in reverse engineering of surface parts is an urgent problem in this field.In addition,as a special type of surface,quadric surface(sphere,conic surface and paraboloid surface,etc.)is widely used in industrial machinery and equipment.Because of the uniform expression of quadric surface,it is different from that of surface in the process of surface information generation.Therefore,this paper studies the surface information generation method of reverse engineering of surface and quadric surface parts to improve the precision and efficiency of surface information generation of reverse engineering of mechanical parts composed of surface or quadric surface.The main research work of this paper is as follows:Firstly,the precision of surface information generation in reverse engineering of surface parts is studied.The surface point cloud data is processed to obtain the elevation image of the surface.Then the discrete stationary wavelet transform is used to transform the discrete elevation sequence corresponding to the surface elevation image,and the corresponding wavelet detail coefficient is obtained.The location of the impact signal of the wavelet detail coefficient is extracted by the difference processing of the adjacent wavelet detail coefficient,and then the feature points of the surface of the surface are extracted.The data points needed for NURBS surface fitting are composed of feature points and selected space points between feature points.The control points were calculated and NURBS surface fitting was performed to improve the precision of surface information generation in reverse engineering of surface parts.Secondly,the generation efficiency of surface information in reverse engineering of surface parts is studied.The BP neural network is used to train the node vectors required for NURBS surface fitting,and the nonlinear mapping relationship is established.The prediction model of node vector based on BP neural network is established.On the basis of this prediction model,particle swarm optimization algorithm is added to optimize the BP neural network node vector,accelerate the convergence speed,realize the rapid prediction of node vector.The node vectors predicted by the optimized node vector prediction model were used in NURBS surface fitting to improve the generation efficiency of surface information in reverse engineering of surface parts.Thirdly,the precision of surface information generation in reverse engineering of quadric surface parts is studied.The mesh model is obtained by using Delaunay mesh method.The least square fitting method is used to calculate the nearest neighbor plane and its normal vector of each spatial point of point cloud.According to the mapping relationship between point cloud data and normal vector,the characteristic parameters of the quadric surface can be calculated to improve the generation accuracy of reverse engineering surface information of parts composed of quadric surface.Finally,the surface information generation efficiency of quadric surface parts reverse engineering is studied.The linear interpolation algorithm is used to process the point cloud data of the quadric surface,and the interpolation model is obtained.The characteristic equations and points are obtained by using fractional Fourier transform.The boundary points of quadric surface are extracted by combining the location information of spatial points in quadric surface point cloud data.The corresponding equation of quadric surface is constructed,the corresponding point cloud data is simulated,and the corresponding characteristic coefficients are obtained.The correlation coefficient method is used to calculate the similarity between the original feature coefficient and the corresponding feature coefficient,and identify the type and feature information of the quadric surface under test,so as to improve the generation efficiency of the surface information of the reverse engineering parts composed of quadric surface.