Robust Parameter Design With Feedback Control

Robust parameter design (or, briefly, parameter design) has been widely used as a cost effective tool to reduce process variability by appropriate selection of control factors to make the process insensitive to noise. However, when there are strong noise factors in the process, use of parameter design alone may not be effective and an on-line control strategy can be used to compensate for the effect of noise. In this paper, a parameter design methodology in the presence of feedback control is developed. Systems with pure-gain dynamic model are considered and the best proportional-integral (PI) and minimum mean squared error (MMSE) control strategies are developed by using parameter design. The proposed method is illustrated using a real life example from a urea packing plant.