论文部分内容阅读
The existing research of the flatness control for strip cold rolling mainly focuses on the calculation of the optimum adjustment of individual flatness actuator in accordance with the flatness deviation , which is used for general flatness control.As the basis of flatness control system , the efficiencies of flatness actuators provide a quantitative description to the law of flatness control.Therefore , the determination of actuator efficiency factors is crucial in flatness control.The strategies of closed loop feedback flatness control and rolling force feed-forward control were established respectively based on actuator efficiency factors.For the purpose of obtaining accurate efficiency factors matrixes of flatness actuators , a self-learning model of actuator efficiency factors was established.The precision of actuator efficiency factors can be improved continuously by the input of correlative measured flatness data.Meanwhile , the self-learning model of actuator efficiency factors permits the application of this flatness control for all possible types of actuators and every stand type.The application results show that the self-learning model is capable of obtaining good flatness.
The existing research of the flatness control for strip cold rolling mainly focuses on the calculation of the optimum adjustment of individual flatness actuator in accordance with the flatness deviation, which is used for general flatness control. As the basis of flatness control system, the efficiencies of flatness actuators provide a quantitative description to the law of flatness control. Before, the determination of actuator efficiency factors is crucial in flatness control. The strategies of closed loop feedback flatness control and rolling force feed-forward control were established respectively based on actuator efficiency factors . For the purpose of obtaining accurate efficiency factors matrixes of flatness actuators, a self-learning model of actuator efficiency factors was established. Precision of actuator efficiency factors can be improved continuously by the input of correlative measured flatness data. Meanwhile, the self- learning model of actuator efficiency factors permits the application of this flatness control for all possible types of actuators and every stand type. The application results show that the self-learning model is capable of obtaining good flatness.