Sensor Fusion Helps to Improve Strip Speed Measurement in Cold Rolling Mills

Abstract

Reliable and accurate measurement of strip input and output speeds is essential for several types of control algorithms in cold rolling mills. The combination of polynomial regression and subsequent sensor fusion can improve signal quality when it is possible to employ two speed measurement principles simultaneously. The developed algorithm, which aims to minimize the time delay inherent in conventional digital filters, is designed for online use. The data from two dissimilar rolling mills were used to test and validate the method. Issues of practical implementation of the algorithm are discussed