Predictive control of a solution copolymerization reactor using locally linear identifier and evolutionary programming optimizer

This paper, presents an implementation of a predictive control algorithm using locally linear model tree (LOLIMOT) as the system identification algorithm, on an industry case study: solution copolymerization of methylmethacrylate and vinyl acetate. The nonlinear system is represented with a model based on LOLIMOT algorithm. The control approach is successfully implemented during a transient from one operating point to another. Significant improvement is observed in reducing the transient period. Simulation results conform the effectiveness of the proposed predictive control strategy.