Continuing requirements for smaller flat-rolled product thickness tolerances are resulting in the increasing depence of gage control schemes on the capabilities of hydraulic gap control systems. Automatic gauge control (AGC) systems are defined as the control systems that are responsible for on-line reguration of the mill equipment to obtain the rolled product thickness within the target tolerances. An AGC systems consists of a number of control systems that directly or indirectly affect the roll material thickness. The three principal control systems that are considered as the main components of an AGC system are: a) Roll gap control. b) Mill speed control. c) Strip tension control.
In this work, the roll gap control in AGC systems is focused that is constructed as the hydraulic servo system. Nowadays, as the conventional controller for the roll gap control the P controller is used. But, the hot-rolling process includes highly nonlinear components on the characteristics of the process. These system nonlineality affects seriously thickness accuracy in the mill stand with linear controller as P controller. In the present work, two controllers are proposed for the hydraulic gap control. These are linear controller and nonlinear controller. As linear controller, PD controller is proposed due to better frequency responses than the conventional P controller for the thickness variation of rolled material. As nonlinear controller, TDC (time-delay control) controller is proposed against nonlinearity of the process. For experimental studies of hydraulic gap control system in a single stand of a typical hot strip mill, Model-tester is designed on a mill builder's premises.
On this Model-tester, the comparison of proposed controllers is executed. For the application of TDC controller to Model-tester, the stability of internal dynamics in Model-tester is shown. The TDC controller is designed to ensure the asymptotic stability of the closed-loop system. For the comparison-experiment, each controller needs to be optimized as a fitness function in a given control gain region. ES (Evolution Strategies) Algorithm is used for optimization of control gain.
Three parts of experiments for the comparison of proposed controllers are: 1) trajectory following test for variation of reference roll position due to the thickness variation of rolled material. 2) robustness test of each controller for material stiffness variation of rolled material. 3) robustness test for various disturbance given during the roll gap control.
Finally, these experiments are performed to show the characteristics of frequency response and robust tracking performance for Model-tester with each controller. Based on valid experimental results, it is shown that nonlinear controller is more applicable to hydraulic roll gap control for thickness accuracy of strip.