Neural control of a steel rolling mill

The authors apply nonlinear neural control to strip thickness control in a steel rolling mill, Different control structures based on neural models of the simulated plant are proposed. The results for the neural controllers, which include internal model control and model predictive control, are compared with the performance of a conventional PI controller. By exploiting the advantage of nonlinear modeling, all neural approaches increase the control precision. The combination of a neural model as a feedforward controller with a feedback controller of the integral type gives the best results