Adaptive calibration of surrogate measurements and its application to the control of moisture content in paper manufacturing

Many industrial processes are in cascade configuration in which the material being processed goes through a sequence of processing units. In many cases, direct in-process measurements of the relevant variable are not available at all but at the last processing unit surrogate or model based soft measurements in which the relevant in-process variables are inferred from other measurements would be useful in these situations. Models necessary for the surrogate measurements however, are often not accurately known. In this paper, we propose a control method in which the surrogate measurement models are adaptively calibrated and the surrogate measurements are used for in-process feedback. The control law is developed in the context of moisture content control for paper manufacturing via successive vacuum dewatering. A Lyapunov based algorithm is derived and proved. Simulation results show that the proposed control strategy can regulate the exiting moisture content of each box at the desired value.