A Novel Process Control for Injection Molding Based Upon On-Line CAE Systems

Abstract

A novel process control system for injection molding based on the integration of Computer Aided Engineering systems (CAE) and process control models has been studied and verified. The objective is to develop a scientific approach for on-line process control by making use of the results from the CAE systems. Initially, the design of injection molded products should be finished with the help of the CAE systems after the dimensions of the products have been determined. From a set of process conditions given by the CAE results, an automatic design of experimentation procedure was conducted for locating the processing window. The processing window was then employed as the CAE inputs for establishing a process model by employing adaptive fuzzy-neural networks. In addition, a set of quasi-optimal process condition was achieved if an optimization objective had been determined based on production requirement. Finally, appropriate cycle-to-cycle control actions on tuning process conditions were taken while the pertinent feedback signals from the process were stored. This process control methodology has been demonstrated with test cases showing promising results.