ATKB-PID: An adaptive control method for micro tension under complex hot rolling conditions

Abstract

At present, the controller parameters in hot rolling roughing microtension control cannot be adaptively adjusted according to the changes of working conditions, which affects the width accuracy as well as the stability of production. Therefore, an ATKB-PID adaptive micro tension control method is proposed in this paper. A linear attention layer is added to the BP-PID controller and a KNN method is established to predict the optimal learning rate and inertia coefficient under actual working conditions. In addition, the objective function is designed according to the production index to improve the performance of the model. Through the comparison experiments with the widely used RBF-BP control and BP-PID control, it is verified that the ATKB-PID control method proposed in this paper has small steady state error and short adjustment time. The ATKB-PID control method can meet the complex and variable micro tension control conditions in thermal roughing, and has good application prospects.