This thesis addresses the hot strip rolling industrial process, where steel strips are deformed through consecutive rolling mill stands to achieve a specific target thickness. The work aims to analyze current control methods and explore new strategies for enhancing the accuracy and uniformity of thickness profiles in finished products. A comprehensive mathematical model of the mill stand is developed and validated, enabling the simulation of the exit thickness profile and roll force based on the strip's entry properties and mill stand inputs. The discussion includes objectives and possibilities for thickness control in hot strip rolling, detailing available actuators, measurements, and typical disturbances. The standard AGC/HGC control structure serves as a benchmark for the newly developed feedforward control concepts, which rely on a static mathematical model of the exit thickness profile. Three feedforward control approaches are examined: one for average thickness in the lateral direction, another for lateral asymmetries, and a third for the entire shape of the exit thickness profile, utilizing an optimization-based algorithm. These control concepts are validated and compared through simulations, with an adaptive strategy that incorporates online estimation of model parameters. The proposed control methods have been implemented in pilot installations and are now in long-term operational use.
