Parameter optimization method of injection molding process based on genetic algorithm and support vector machine regression

To enhance the quality of injection molded products and refine the injection molding process parameters, a hybrid algorithm combining Genetic Algorithm (GA), Support Vector Regression (SVR), and Particle Swarm Optimization (PSO) was introduced. In this optimization framework, PSO-SVR established a Non-linear functional relationship between warpage deformation and key process parameters, including mold temperature, melt temperature, injection time, holding pressure, and holding time (as input variables), with warpage deformation serving as the output variable. A comparative analysis with other prevalent optimization algorithms highlighted the effectiveness of the GA-SVR-PSO hybrid approach in optimizing these parameters. Notably, the PSO-SVR prediction model exhibited superior predictive capabilities, achieving a mean square error of 0.082701, outperforming other models. By employing SVR-PSO as the fitness function and utilizing GA for further optimization, the warpage deformation in injection molded products was reduced by 24.5%, thereby significantly elevating the overall product quality.