Online Monitoring System Design for Roll Eccentricity in Rolling Mills

This paper deals with the parameter estimation and online monitoring of roll eccentricity in rolling mills. Roll-eccentricity-induced disturbances can result in strip thickness deviation and product quality degradation, and a conventional control strategy cannot regulate this kind of periodic disturbances due to its complex characteristics. For the purpose of product quality assessment and monitoring, a performance indicator based on product expectations or the empirical value of quality-related process variables, which can be quantitatively visualized with four zones, is proposed in the high level. In the low level, the key parameters of roll eccentricity are estimated in real time using an adaptive observer for frequency estimation and an adaptive algorithm for amplitude and phase estimation. The performance and effectiveness of the proposed eccentricity monitoring system are demonstrated through a case study of a cold rolling mill from industrial fields.