A Self-Tuning Controller for Head and Tail End Gauge Control in Reversing Cold Mills

摘要

This paper investigates a method to improve theaccuracy of head and tail strip thickness in a reversing mill. Anintelligent PID control approach is applied to the AutomaticGauge Control (AGC) system of the reversing mill. In thisapproach, a neural network is used to tune the parameters ofthe PID controller to improve the accuracy of the exit stripthickness. The effectiveness of the approach has beendemonstrated through a simulation example. The results of thesimulation show that a neural network can reduce output errorquickly and improve the accuracy of the head-end stripthickness effectively while the P1 controller parameters arebeing tuned on-line.