Optimization Of Stamping Process Parameters Based On Parallel BP Neural Network And Its Software Development

As an indispensable manufacturing technology,stamping forming has been widely used in important fields such as automobile industry,chemical industry and national defense.Especially in the automotive industry,automotive stampings are the main component of the car body.If the stamping forming parameter setting is not reasonable,it will cause quality problems such as cracking and wrinkling of the stamping parts,which will ultimately seriously affect the quality and appearance of the vehicle.In the traditional stamping process,the forming parameters are obtained according to empirical formulas or repeated debugging,which has certain blindness.With the substantial improvement of computer performance,the rapid development of surrogate model and finite element technology,the surrogate model combined with finite element method has gradually become a research hotspot in the field of stamping forming parameter optimization.This paper proposes an optimization design method based on parallel BP neural network combined with multi-objective particle swarm optimization to solve the problem of reasonable setting of the stamping parameters.The goal of the proposal is to improve the accuracy of the surrogate model and obtain the optimal stamping forming parameters.The main contents of the paper are as follows:A parallel architecture neural network is proposed to improve the poor generalization performance of BP neural network.By comparing the two neural networks through the test function,the parallel architecture neural network not only improves the poor generalization performance of BP neural network,but also improves the accuracy of the surrogate model.Comparing the neural network with several other surrogate models in different dimensions,it can be seen that this model has certain advantages with the dimensions increasing.In addition,the specific process of the neural network combined with the multi-objective particle swarm optimization algorithm used in the optimization of stamping forming parameters is described.Summarizing three commonly used forming quality evaluation standards,and a novel stamping quality evaluation criteria based on forming limit diagram is proposed.The proposed criteria is more intuitive and the principle is easy to understand.Separate the working area of the forming limit diagram from the non-working area,and only keep the working area of the forming limit diagram to reduce the influence of the non-working area.Then count the number of pixels of different colors in the work area,and apply the number of pixels of different colors as the objective variables of the surrogate model.The optimization method of stamping parameters proposed in the paper is applied to engineering examples.The complexity of engineering examples is different and has different representations.The results show that the optimization method can effectively obtain the optimal stamping parameters.In addition,the relevant software development is displayed,providing the operator with a convenient and simple operating platform.