Study On The Multi-optimization For Process Parameters Of Sheet Metal Forming

Sheet metal forming,a crucial part of modern manufacturing,has been widely applied in vehicle,aviation and military industries.Unreasonable design variables will cause the forming defects such as wrinkle and fracture frequently occurred on work-piece.Meanwhile,sheet metal forming process have the high-nonlinear features that geometric nonlinearity,material nonlinearity and boundary nonlinearity.With the development of numerical simulation technology,in order to reduce the dependence of engineer experience,experimental design method,finite element numerical simulation and optimal method been integrated to optimize the process parameters of sheet metal forming domain.The proposed method can control the forming quality of sheet metal forming process with the low case condition.In this study,central composite design,Gaussian process regression and Multiobjective genetic algorithm been integrated to optimize the parameters of metal forming process.The core of this study are the conducting process of surrogate model and the solving process of surrogate model.A multi-objective optimization method for parameters of sheet metal forming process been established in this paper.Two auto-body drawing case was demonstrated to verify the effectiveness and reliability of proposed approach.The commonly used experimental design methods and surrogate model technologies was compared in this paper,firstly.Secondly,the modeling principle and hyperparameters selection of Gaussian process regression been discussed.Then,based the analysis result of defect mechanism of sheet metal forming process,the selection of design variables been discussed.Meanwhile,based on Gaussian process regression,optimal support vector machine and response surface method,nonlinear regression model between forming process parameters and defect evaluation function is established.After that,compared the regression performance in three regression models to verify the effectiveness and applicability of Gaussian process regression in sheet metal forming process.With the summarizing of the contents,a multi-objective optimization method for sheet metal forming process parameters is proposed.The Pareto based immune genetic algorithm be used to solve the regression model to get the Pareto set of process parameters.The European norm ideal point method proposed to determine the optimal combination of process parameters conveniently with more feasible solution.Then,the finite element analysis and engineering test used to verify the proposed method.Finally,a dash-board drawing case demonstrates that the proposed approach is more accurate and effective than conventional response surface method and traditional finite element analysis method.