Shortcut Design of Divided Wall Columns and Optimization Using Simulated Annealing Algorithm

• Main Authors: Pin-Hsun Huang, 黃品勛
• Other Authors: Jeffrey D. Ward
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/y54kcz

碩士 === 國立臺灣大學 === 化學工程學研究所 === 107 === Divided wall columns (DWCs) are complicated systems for separating multicomponent mixtures. The design of DWCs is more difficult than traditional distillation columns because there are many degrees of freedom. Previous methods developed for designing DWC based on empirical equations do not always work well. Therefore, a new shortcut design method is developed. Aspen Plus simulation and Window-based interface for connecting Aspen Plus and Matlab are used to calculate a preliminary design of DWCs. After that, a stochastic optimization method, a simulated annealing algorithm, is utilized to optimize the system. This algorithm has the characteristic of being a probabilistic technique that can slowly decrease of the probability of accepting worse solution and search for the global optimum solution. This method is applied to optimization problems of two DWC systems and the three types of DWC systems are compared to find which of it is the most economical process.