Injection Molding Parameters Optimization through a Hybrid System of Artificial Neural Network and Genetic Algorithm

Nowadays competitive conditions force us to faster and cheaper production with a higher quality. The use of Computer-aided analysis and engineering softwares such as MoldFlow Plastic Insight (MPI) could help engineers to have initial knowledge about the plastic injection processes such as injection, packing, cooling, ejection and process/part quality control that will be undertaken for the parts, which are designed to beproduced by plastic injection method. In this study, MPI was applied to generate responses such as average volumetric shrinkage (shrinkage) and in-mold pressure (pressure). Process parameters such as mold temperature, melt temperature and gate location, are considered as model variables. The objective of this research is to obtain an optimal process parameters corresponding to minimum shrinkage and pressure. At first Taguchi method is used to solve the minimizing problems, separately. Then two three-layer Back- Propagation (BP) Artificial Neural Networks (ANN) are used to modeling the relationship between processing parameters and part shrinkage and also pressure, separately. A couple of ANN and Genetic Algorithm (GA) is used to solve the two objective problem and to obtain the optimal parameter values and set of model variables leading to minimum shrinkage and pressure. Finally, the optimal set of variables was compared with sets that obtained from Taguchi method analyze for minimum shrinkage and minimum pressure, separately. This compare proves that couple of ANN/GA has reasonable performance and also shows that use of this hybrid method enhances optimization power in optimization of process parameters.