1993 Volume 79 Issue 3 Pages 373-379
This paper describes a new robust method for thickness control of reverse cold rolling mills. Recently, high accuracy in thickness is required exceedingly. However, there are fluctuation of parameters, nonlinearity, lag-times in the inputs, lag-times in the outputs and so on, in the nonstationary state under accelerating and decelerating conditions. They make accuracy in thickness worse, and it is necessary to design control systems that have robustness for them. In this new approach, mutual interaction between thickness control and tension control is decoupled. Padé approximation is employed for the lag-times in the inputs, and the disturbance observer is applied to compensate the fluctuation and the nonlinearity. The lag-times in the outputs are compensated by a predictive control method. Simulation studies show that the new approach improves the response of the thickness control system and achieves the robustness for them, and that the delivery thickness deviation is obviously reduced. This robust control system has been applied to the actual rolling mill, and the strip length without tolerance is remarkably shortened.