Application of a simulated annealing algorithm to design and optimize a pressure-swing distillation process. (5th December 2016)

• Record Type: Journal Article
• Authors: Wang, Yinglong; Bu, Guangle; Wang, Yongkun; Zhao, Tingran; Zhang, Zhen; Zhu, Zhaoyou
• Is Part Of: Computers & chemical engineering. Volume 95(2016)
• Page Start: 97
• Page End: 107
• Publication Date: 2016-12-05
• Subjects: Pressure-swing distillation -- Simulated annealing algorithm -- Optimization method -- Azeotropic mixtures; Chemical engineering -- Data processing -- Periodicals; 660.0285
• DOI: 10.1016/j.compchemeng.2016.09.014
• Languages: English
• ISSNs: 0098-1354

Abstract

Highlights: Pressure-swing distillation process is optimized by simulated annealing method. Judgment modules and parameter settings of simulated annealing are investigated. The proposed method can optimize process with no, partial and full heat integration. This method has high probability to obtain the global optimum. This method achieves automatic optimization with less computing time. Abstract: The design and optimization of pressure-swing distillation (PSD) have a critical impact on its economics. An optimization method based on simulated annealing algorithm (SAA) was proposed. The move generator and cooling schedule of the SAA were discussed, and suitable parameter settings were investigated. Two cases of PSD with and without heat integration were optimized by the SAA-based optimization method using procedures of pressure specified and pressure optimized. The results of the process without heat integration were compared with conventional optimization methods. For the acetone-methanol system, the total annual cost (TAC) shows a 5.69% decrease with the pressure specified and a 17.32% decrease with the pressure optimized. For the methanol-chloroform system, the TAC shows a 1.79% decrease with the pressure specified and a 9.04% decrease with the pressure optimized. The SAA-based optimization method has the advantages of a high probability to obtain the global optimum, automatic calculation, and less computing time.