Optimal grade transition of continuous polymerization process based on multiobjective particle swarm optimization

Abstract

This paper presents an approach that combines the linear decreasing weight particle swarm optimization (LDWPSO) algorithm with crowding distance diversity maintenance to solve the multiobjective optimization problem (LDW-MPSO) for the optimal grade transition of continuous polymerization process. In the case study of a free radical styrene polymerization in a pilot continuous stirred tank reactor (CSTR), the grade transition optimization problem is described using two objective functions. The first focuses on the polymer quality, and the second aims to minimize the raw material consumption during the grade transition process. The optimal trajectories of manipulated variables, operating conditions and polymer quality are obtained by solving the multiobjective optimization problem using LDW-MPSO. The simulation results confirm the effectiveness of the presented approach to shorten the transition time and minimize the material consumption.