Research on dynamic flatness feedback control strategy based on IGWO control efficiency identification for cold tandem rolling mill

XM Zhou and LQ Li and SK Wang and QX Xiong, INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS, 16, 5397-5417 (2025).

DOI: 10.1007/s13042-025-02577-1

Flatness control is an important means to improve the quality of the strip, and modern tandem cold rolling mills are equipped with various control actuators. The existing flatness control efficiency is determined offline and remains constant throughout the manufacturing process. During actual manufacturing, equipment, and process states may change, making it necessary to identify flatness control efficiency online. The main flatness feedback control strategy is a fixed priority sequence in the existing CVC (Continuous Variable Crown) cold tandem rolling mill. The control system does not have the flexibility to adjust the sequence based on real-time flatness. Aiming at the above problems, in order to improve the effect of flatness control, this paper proposes a dynamic flatness feedback control strategy based on control efficiency identification. To address the issue of high cost and difficulty in updating the on-site flatness control efficiency, based on historical actual data, the IGWO (improved grey wolf optimization) algorithm is used to intelligently identify control efficiency. Data post-processing through IF (isolated forests) and CLT (central limit theorem) obtains results on control efficiency. At the same time, a similarity index is proposed to dynamically adjust the priority sequence of feedback control based on the identification results of the control efficiency and the measured flatness deviations. Finally, the field application results show that the flatness control accuracy is improved, the reduction value of the flatness deviations increased by 13.41%, and the reduction rate of the flatness deviations increased by 9.66%.

Return to Publications page