Application of Generalized Minimum Variance Algorithm to Polymerization Reactor at a Constant Set Point

Abstract

Generalized Minimum Variance Control (GMV) has been applied to a jacketed batch reactor to control the temperature of a free radical solution polymerization of styrene and its performance was examined. As controller, an online digital computer was used to keep the reactor temperature at pre-calculated optimal set point. A Pseudo Random Binary Sequence (PRBS) signal and Bierman algorithm were used to estimate the Auto Regressive Integrated Moving Average (CARMA) model parameters for GMV control system. Heat input to the polymerization reactor was chosen as a manipulated variable. Performance results of GMV were compared experimentally and theoretically and with the results obtained by using conventional deadbeat algorithm. GMV control system shows satisfactory behavior to track the reactor temperature.