Incorporation of phenomenological models in a hybrid neural network for quality control of injection molding

Injection molding is characterized by complex dynamics, which makes quality difficult to control. This is because the exact relations among the machin e inputs, material properties, and molded part quality are not known precis ely. Hence, the existing models for quality prediction have a limited accur acy and difficulty in application to general molding applications. This art icle investigates the integration of analytical process knowledge and artif icial neural networks as a solution for quality prediction of molded parts, with accuracy increased toward quality control targets of three defects pe r million (6 sigma), This article describes the hybrid system based on the neural network and process knowledge, then compares its performances with c onventional neural models for the prediction of the injection pressure.