Model predictive control based on a detailed model and observer for strip steering control in a hot strip finishing mill

Abstract

In a hot strip finishing mill, an unstable lateral steel strip movement called strip walking causes the strip to crash into side guides placed at the entrance of the stand mill, resulting in a tail pinch where the strip is pressed into a folded state. Therefore, strip walking should appropriately be suppressed to prevent tail pinches, requiring adequate strip steering control. To this end, various strip steering control methods have been proposed. Among them, model predictive control, which explicitly considers the motion of a steel strip and constraints on the actuator’s speed, is expected to be of great practical use. However, the disadvantages of this control are that the model’s accuracy affects control results, state variables must be estimated, and its optimization computation is time-consuming. Hence, this paper proposes a method to overcome such disadvantages by adopting a model describing adequately strip walking, model predictive control using discretized inputs, and a deadbeat current observer to estimate state variables. Subsequently, we demonstrated the effectiveness of the proposed method using several numerical simulations together with setting discrete values for the discretized inputs and setting weights in the cost function to be minimized in the model predictive control. In conclusion, it is clarified here that the selection of a set of discretized control inputs, which maximize the leveling speed in minimum time and the proper selection of weights in the evaluation function are crucial for the practical use of model predictive control to prevent strip walking.

Keywords

• steel industry
• hot strip rolling process
• strip walking model
• modeling
• model predictive control
• strip steering control
• tail pinch