Optimal design for dividing wall column using support vector machine and particle swarm optimization

Abstract

Dividing wall column (DWC) is a practical method to reduce energy cost and save capital cost in process intensification technologies for distillation columns. However, the optimization of DWC is a highly nonlinear and multivariable problem, which requires intensive research and investigation. The purpose of this paper is to study the optimal design for DWC using a combination of support vector machine (SVM) and particle swarm optimization (PSO). By generating a SVM model for optimization of DWC, the speed of calculation improves a lot. And SVM is based on rigorous mathematical theory, which overcomes the effect of human experience. PSO is applied to optimize the thermally coupled liquid and vapor flow rate and the stage numbers of different DWC sections. The proposed combination method achieves an encouraging result and proves to be a promising method for optimal design of DWC.