Effective Method to Compensate the Nonlinear and Time-delay Part of AGC Strip Gauge Control

Against the problems of time-delay, parameters-varyingness, large inertia, and non-linearity existing in AGC systems, a controller which combines adaptive centralized time-delay neural network recognition, optimal prediction and immunized fuzzy PID is proposed. The controller effectively improves the accuracy and dynamic performances of the system. Simulations in Matlab show that the controller is superior to traditional control strategies in all design criteria.