Optimization Of Automobile Covering Parts Forming Based On Grey System Theory And Neural Network Genetic Function Optimization

The stamping process of automobile covers are complex process of elastic-plastic and large deformation for the shape of complex and high precision.With the rapid development of numerical simulation technology,a large number of scholars have been using numerical simulation technology to predict the possible problems of sheet metal forming process.Using the technology can not only reduce mold testing times,but also optimize the structure of the mold design and process design.However,due to the complex structure of the automobile panel,it has high quality requirements,simply using numerical simulation technology has a high blindness.Therefore,although taking a lot of time to do simulation,the experimental results are still unsatisfactory.How to make good use of various optimization algorithms and combine various algorithms to adjust the forming process parameters in the stamping process so as to obtain the combination of process parameters to meet the quality requirements have always been a hot issue and difficult problem in the automobile manufacturing field.Based on GS theory and neural network genetic algorithm function optimization method,this paper use the nonlinear finite element analysis software Dynaform to analyze the stamping process of left side rear panels,and to optimize process parameters,which manufactured by FAW-GM Automobile Manufacturing Co.,Ltd.Firstly,based on the GS theory,the gray correlation analysis of the acquired data is carried out by using the orthogonal test method,and then the maximum reduction rate is obtained under the combination of different process parameters.The two most important factors of the rate are the stamping speed and the blank holder force.Moreover,using the Latin hypercube sampling method to randomly select the two main factors within a given range,and then obtaining the maximum reduction rate by using non-linear finite element software Dynaform to analyze.Secondly,based on the neural network genetic algorithm function optimization model,the Latin hypercube sampling is used to obtain the stamping speed and the blank holder force in the data as the input and the maximum thinning rate as the output,and the input and output data are used to train the BP neural network.The genetic algorithm is used to predict the BP neural network after the training as the individual fitness value,that is,the result of minimizing the maximum reduction rate is obtained and the corresponding input value is obtained.Lastly,Dynaform software is used to simulate the optimal combination of process parameters,and the experimental platform is used to verify the experiment.The numerical simulation results and experimental results show that the optimized stamping parameters can improve the forming performance of the sheet.This method can effectively predict and optimize the process parameters of automobile panel.