Injection molding process optimization of plastic packaging device based on GA and BP neural network

This paper investigates a plastic sealing device layering problem, A BP neural network combined with genetic algorithm (GA) is proposed to optimize injection molding process parameters based on Moldex3d mold flow analysis. Firstly, an orthogonal test is established, with five process parameters, namely mold temperature, injection pressure, holding pressure, holding time and injection speed, as the factor variables and warpage deformation amount as the optimization index, and the initial parameters are obtained by using Moldex3d software for mold flow analysis. Then the BP neural network is established by using Matlab software to train the parameters and obtain the objective function. The total warpage of the plastic device is significantly reduced, the delamination problem of the plastic device is improved to a certain extent, and the optimal combination of process parameters is obtained. The warpage amount is smaller than the orthogonal test results, which is verified by Moldex3d simulation. The optimization of parameters and improvement of delamination problem proved the feasibility of the injection molding process optimization approach of plastic sealed devices based on Moldex3d, GA and BP neural network.