This paper presents a new approach for temperature control of an injection molding machine (IMM) that uses a model predictive control (MPC) strategy. The control system consists of a number of single-input-single-output model predictive controllers, each associated with a particular temperature zone. What distinguishes this approach from others is how the MPC strategy exploits knowledge of temperature interaction between adjacent zones and the effects of back pressure, to develop individual temperature controllers for each zone. This is achieved by decoupling the interaction between zones. The new thermal controller was simulated and implemented with good results on a 150-tonne IMM using a series of comparative experiments.

This paper develops a mathematical model for the dynamics of a plastic injection molding machine (IMM) that may be used for the design of a temperature-control system. The research in this paper is novel in comparison to others since the derived models explicitly include the effects of zone interaction and backpressure, and do not lump these into an arbitrary disturbance signal. A series of experiments were conducted on a 150-tonne IMM to identify the parameters of the proposed model using measurements of zone temperatures, percentage heater input, backpressure and screw speed. The identified model was validated using a series of blind tests that compared the model output with the measured barrel temperatures of the IMM.