Multi-objective Optimization of Injection Molding Process Based on One-Dimensional Convolutional Neural Network and the Non-dominated Sorting Genetic Algorithm II

Abstract

In the process of injection molding, the vacuum pump rear housing is prone to warping deformation and volume shrinkage, which affects its sealing performance. The main reason is the improper control of the injection process and the large flat structure of the vacuum pump rear housing, which does not meet its production and assembly requirements (the warpage deformation should be controlled within 1.1 mm and the volume shrinkage within 10%). To address this issue, this study initially utilized orthogonal experiments to obtain training samples and conducted a preliminary analysis using gray relational analysis. Subsequently, a predictive model was established based on a one-dimensional convolutional neural network (1D CNN). Input parameters from the injection molding process, including melt temperature, mold temperature, packing pressure, packing time, injection pressure, injection time, and cooling time, were used while warping deformation and volume shrinkage were considered as outputs. Global optimization was performed using the non-dominated sorting genetic algorithm II (NSGA-II), and the optimal combination of process parameters was evaluated using the criterion importance through intercriteria correlation—technique for order preference by similarity to ideal solution (CRITIC-TOPSIS). Moldflow analysis demonstrated that the obtained indicators outperformed the optimization results from orthogonal experiments, confirming the effectiveness of the injection molding process parameter optimization method based on 1D CNN-NSGA-II. In comparison to the pre-optimization results, product warping deformation decreased by 40.68%, and volume shrinkage reduced by 18.14%, and all of them meet the production requirements.