Intelligence-Mechanism Synergistic Method for Cold-Rolled Strip Deviation Measurement and Flatness Measurement Error Compensation

Contact flatness meters are essential for the online measurement and control of the flatness of cold-rolled strips. However, strip deviation during cold-rolled production can lead to varying degrees of errors in flatness measurement, impacting control accuracy. To address this, this paper proposes an intelligent-mechanism synergistic method for cold-rolled strip deviation measurement and flatness measurement error compensation. First, a high-speed industrial camera is used to capture production data, and perspective distortion caused by spatial inclination is corrected via the perspective transformation principle to construct a dataset. Next, an intelligent strip area segmentation model is built via a fast semantic segmentation algorithm, and an edge position extraction algorithm is designed on the basis of the segmentation results to calculate real-time strip deviation. Finally, taking actual production conditions into account, this study develops a flatness measurement error compensation mechanism model that considers the influence of strip deviation on the edge channel flatness and overall flatness distribution. This method was successfully implemented at a 600 mm cold-rolled copper strip production site. On-site verification demonstrated that the proposed deviation measurement method is both accurate and fast. The compensated online flatness measurements closely match the actual flatness of the strip, making a significant contribution to improving the accuracy of cold-rolled strip flatness control.