Study On Predictive Control Of Monomer Recovery Column In Acrylonitrile Copolymer Process

Monomer recovery distillation column which is used to remove the monomer from the mixture is an important equipment in the process of monomer recovery, time-variable time delay is a big problem as well. Although the traditional control method can achieve the safety aims, it often costs more energy. This research applies the theory of generalized predictive control to achieve the control process of monomer recovery distillation column in order to get the better performance.Monomer recovery distillation column model which is implemented for the column outlet product quality is studied in order to avoid the problem which is generated by regarding the temperature as indirect quality criterion. On the basis of these, the step response value is obtained then the translate functions of the monomer recovery distillation column are achieved by least square algorithm. Appropriate sampling period is chosen to obtain controlled auto-regressive integrated moving-average (CARIMA) model which will be used for the generalized predictive control.Generalized predictive control with direct quality criterion based on the CARIMA model is adopted in order to realize a better operating mode. Implicit generalized predictive control algorithm which recognizes the controller parameter directly using the recursion least squares method is researched to avoid solving the diophantine equation and the matrix inverse. This article analyzed the influence of main parameters of the generalized predictive control algorithm. In view of the characteristic of monomer recovery distillation column, appropriate controller parameter has been chosen. When monomer recovery distillation column model has time-variable time delay, this research compared implicit generalized predictive control algorithm with Smith predictor algorithm in order to present the evidence of the validity of implicit generalized predictive control algorithm.