Design of performance-adaptive PID controllers

In the challenge to manufacture high quality products for less, it is necessary to regularly monitor performance of control loops that regulate the quality variables of interest. This paper describes performance-adaptive PID control schemes which is based on a unified approach to the design of a control performance and PID controller. According to the proposed approach, the control performance is first monitored regularly. Then, if the performance exceeds a user-defined threshold, the system identification is initiated and PID parameters are subsequently updated for the new model. In this paper, two performance-adaptive PID controller design schemes are introduced. One is that the control performance is evaluated based on the minimum variance index of the control error, and PID parameters are returned based on the relationship of the generalized predictive control. Another is that the modeling performance is first evaluated, and PID parameters are subsequently calculated based on the LQG trade-off curve obtained for the re-identified process model. The behavior of the performance-adaptive PID control schemes is numerically and experimentally evaluated.