Multistage dynamic optimization of a copolymerization reactor using differential evolution

Abstract

A multistage dynamic optimization methodology with sequential implementation procedure is proposed, and the evolutionary optimizing features of differential evolution (DE) are exploited to implement the methodology for optimal control of a semi-batch copolymerization reactor. DE is designed and implemented to determine the optimal control policies for monomer addition rate and reactor temperature to produce a polymer with the desired copolymer composition and molecular weight distribution. Further, a similar multistage dynamic optimization strategy based on iterative dynamic programming is used for optimal control of copolymerization reactor and to compare with DE. The results show the effectiveness of the DE-based multistage dynamic optimization strategy in determining the optimal control policies that yield the desired polymer product characteristics. © 2013 Curtin University of Technology and John Wiley & Sons, Ltd.