Model predictive control helps to regulate slow processes--robust barrel temperature control.

Slow temperature control is a challenging control problem. The problem becomes even more challenging when multiple zones are involved, such as in barrel temperature control for extruders. Often, strict closed-loop performance requirements (such as fast startup with no overshoot and maintaining tight temperature control during production) are given for such applications. When characteristics of the system are examined, it becomes clear that a commonly used proportional plus integral plus derivative (PID) controller cannot meet such performance specifications for this kind of system. The system either will overshoot or not maintain the temperature within the specified range during the production run. In order to achieve the required performance, a control strategy that utilizes techniques such as model predictive control, autotuning, and multiple parameter PID is formulated. This control strategy proves to be very effective in achieving the desired specifications, and is very robust.