| Finishing mill control is vital in meeting the product dimensional quality requirement for hot sheet rolling. In a finishing mill, strip tension is regulated by loopers to ensure a uniform strip width. Various modern control approaches have been applied so far to improve looper control, including: optimal, inverse linear quadratic, H-infinity and robust control. In this study, a computer simulation of a finishing mill has been developed to design and evaluate new control strategies. The model has been validated against real plant data and can predict the looper motion and phenomena such as deformation, heat transfer and slip. It is nonlinear and accounts for the variable inter-stand transport delay. Looper covariance controllers have been designed and tested using the simulator. It has been found that covariance control provides a more stable operation compared to the conventional control, and use of speed disturbance information in feedforward mode significantly enhances any form of looper control. |
