Identifications of quality parameters in paper-making industry

Paper basis weight, moisture and ash are the important specifications of paper quality. The reason of the difficulties to be controlled well lies on that a lot of factors have influences on them. The work focuses on the multivariable and closed-loop identification of paper-making industrial processes, which is used in model predictive control strategy. The brief arts and crafts of the ordinary fourdrinier are first described in this paper. The motivations and advantages for closed-loop system identification are also given. After analysis of the intractable points of paper basis weight, moisture and ash control, a plant is identified both from open-loop and closed-loop data. The model obtained from the open-loop data is used as reference to evaluate a closed-loop system identification approach. Through different noise-to-signal ratios, the paper gives a conclusion that pre-filtration can improve the precision of closed-loop identification model. Proposition of a novel criterion to evaluate the identification model is another conclusion.