An Application of Stochastic and Dynamic Models for the Control of a Papermaking Process

A statistical approach to model a papermaking process and to control paper basis weight is presented. The process behavior is described by a mathematical model which takes into consideration both the input-output dynamics and disturbances coming into the system. The model building procedure was based on a three-step approach of identification, estimation and diagnostic checking. The disturbance model is accounted for by a discrete time series model, whereas the dynamic model is a discrete transfer function model. The simplicity of the modeling process lies in the fact that only the input-outpul data are necessary without recourse to a complicated analysis of the physical system itself. An optimal control equation was arrived at whereby the basis weight of the paper was to be controlled at a specified target value. Confirmation runs were conducted to check the effectiveness of the control equation.