Optimization Design Of Auto-body Sheet Metal Location Layout Based On Prediction Model

The auto-body is assembled by hundreds of flexible sheet metal parts at different workstations.There are various factors in the assembly process,such as parts manufacturing deviation,sheet metal positioning deviation,welding springback and clamping deformation.These factors are coupled layer by layer and spread and amplified step by step,seriously affect the auto-body dimensional quality.The sheet metal positioning deviation is one of the key factors affecting the auto-body dimensional quality.In recent years,domestic and foreign scholars mainly focus on the optimization of location point or hole position in the research of location layout of auto-body sheet metal,and lack of research on joint optimization of location point and hole position.In addition,the existing design of sheet metal location layout mainly relies on human experience and lacks the guidance of systematic theoretical methods.At the same time,the finite element method is used to analyze the deformation of sheet metal,although the calculation accuracy is high,but in the calculation of complex assembly,the calculation efficiency is low and can not meet the needs of actual production.In order to reduce the design changes,shorten the development cycle and improve the auto-body manufacturing quality.Firstly,this paper summarizes the deviation transfer model of fixture location layout and the mathematical expression of stability parameters.On this basis,based on multi-objective genetic algorithm,the coordinates of the first three location points in the main plane of sheet metal "3-2-1" positioning are designed.Then,on the basis of finite element analysis,a radial basis function neural network prediction model is constructed,and its prediction results are taken as individual fitness values,and the position of the fourth location point is determined by particle swarm optimization algorithm.Finally,the deformation of the sheet metal under the constraint of the location point is considered in the optimization of the coordinate position of the locating hole,taking the deviation of the key measuring point of the sheet metal as the target,establishing response surface prediction model to optimize locating hole coordinate position,the B-pillar reinforcement plate is taken as an example to apply the above research.The main contents of this paper are as follows:First of all,based on multi-objective genetic algorithm,the layout of three locating points of the first datum plane in sheet metal "3-2-1" positioning is designed.On the basis of rigid body model,the mathematical models of deviation transfer and the mathematical expressions of stability parameters in existing literatures are summarized.The traditional method only considers a single optimization objective(such as deviation transfer path)or transforms multiple optimization objectives into a single objective(weighted coefficients are allocated by manual experience),it did not fundamentally solve the multi-objective optimization problem of sheet metal positioning layout.In order to solve this problem,a multi-objective genetic algorithm is proposed to design the layout of the first three locating points in the main plane of thin plate "3-2-1" positioning.Secondly,establishing the prediction model of sheet deformation and determining the optimal location of the fourth location point of the sheet metal under the "4-2-1" positioning layout.On the basis of determining the position of the first three locating points in the main plane,taking the position of the fourth locating point as the input and the average deformation of the sheet metal as the output,a radial basis function neural network prediction model is established,and the predicted results are taken as the individual fitness value.The location of the fourth locating point,which minimizes the average deformation of the sheet metal,is determined by particle swarm optimization.Thirdly,establishing the response surface prediction model to optimize locating hole coordinate position.The coordinate location optimization of sheet metal location holes is studied in depth.Taking the deviation of the key measuring points of the clamped sheet metal as the optimization objective and the coordinates of the locating holes as the design variables,and combining with the 3DCS deviation simulation software,the mathematical model of the coordinate parameters combination of the locating holes of the sheet metal holes is established.The influence of the interaction between different coordinates on the deviation of the measuring points of the sheet metal is analyzed,and the locating hole layout scheme which can effectively reduce the deviation of the measuring points is found through multi-objective optimization.Finally,the B-pillar stiffening plate is used as a validation model to verify the above research content.Based on multi-objective genetic algorithm and radial basis function neural network prediction model,the location of B-pillar stiffening plate is optimized.With the key measuring point deviation of B-pillar stiffening plate in clamping as the optimization objective and the location of hole coordinates as design variables,the response surface prediction model is established to optimize the location of hole coordinates,and is compared with the original design scheme of B-pillar stiffening plate.The results show that the optimized positioning layout can effectively reduce the clamping deformation and assembly deviation of B-pillar stiffening plate.