Control of a Die Casting Simulation using an Industrial Adaptive Model-Based Predictive Controller

摘要

Mathematical modeling and simulation is a common technique to improve and optimize industrial processes. Complex models of casting processes are available to describe the evolution of process variables such as temperature and pressure, and can be used to reliably predict defect formation. Similarly, advances in control theory have led to widespread use of online control for the majority of industrial processes. A multivariable control scheme has been developed to control a 2-D axisymmetric model of an industrial low-pressure wheel die casting process for use in assessing predictive control strategies to reduce defect formation. Temperatures at various locations within the die are controlled to predefined set points based on the cyclic steady state temperature profile. The volume flow rates of cooling media in multiple channels are used as control variables. The incoming molten aluminum temperature and die open time are used as measured disturbance variables for the controller.