Modelling feedback control adjustment to control output product quality

This paper describes a method to model feedback control adjustment to control the quality of products at the output of a continuous process. The method takes into consideration situations in which the product quality mean is on target and not on target by assigning suitable probabilities for the respective situations. A test of significance is conducted for such situations when a feedback control adjustment is required to bring the product quality mean on target/controller setpoint. In making feedback control adjustments, it is probable that because of its automatic nature, deficiencies in the feedback control loop such as inexact feedback error and incomplete feedback compensation, the feedback control adjustment is incomplete. A method is proposed in this paper for a test of significance whether a feedback control adjustment is required or not in a process subject to random shocks at the input. A probability model for feedback control adjustment based on a simple probability test for feedback control adjustment indicates that the product quality mean is not on target by sounding process ´out of control´ alarm signals. Knowing the probability of the required feedback control adjustment, through the probability test, the process can be adjusted for this situation.