Design of fuzzy-PI smith controller for Automatic Gauge Control in cold rolling mill

In this paper, a fuzzy-PI smith controller is derived for Automatic Gauge Control in a single-stand cold rolling mill. The controller consists of smith predictive control, fuzzy control, and PI control. The smith predictive control method is proposed to eliminate the time delay generated by the thickness measurement. Furthermore, the fuzzy control scheme is used to tune the parameters of PI controller automatically. The verification of control performance for the fuzzy-PI smith controller is conducted using Hardware-in-The-Loop Simulator (HILS). The experimental results show that the proposed controller yields smaller undershoot, faster settling time, stronger robustness, and higher control accuracy comparing to the conventional PI controller which is used in Monitor AGC.