Model Predictive Synchronous Control of Barrel Temperature for Injection Molding Machine Based on Diagonal Recurrent Neural Networks

Abstract

A nonlinear model predictive control (NMPC) based on diagonal recurrent neural network (DRNN) was used to control multisection barrel melt temperatures of an injection molding machine. In this method a DRNN was used to construct a nonlinear predictive model of barrel melt temperatures and genetic algorithm (GA) was used as a rolling optimization tool. Simulations and experimental results show that this method not only guarantees the accuracy of temperature control of barrel melt temperatures but also improves synchronization of barrel temperature control and it improves the consistency of the barrel melt polymer and the quality of the molded parts.

Keywords:

• Barrel temperature
• Diagonal recurrent neural networks
• Model predictive control
• Synchronous control

ACKNOWLEDGMENTS

This project is sponsored by the Zhejiang Provincial Natural Science Foundation of China (Y1110135) and the Foundation of Zhejiang Educational Committee (Y201009953).