Optimization of heat exchanger network synthesis via evolutionary algorithm

• Main Author: 曾惟哲
• Other Authors: 王逢盛
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/85969737109186307908

碩士 === 國立中正大學 === 化學工程研究所 === 88 === For many plants such as chemical materials, petroleum chemistry, polymer, fertilizer chemistry and fiber chemistry, heat exchangers always play an important role in the industrial processing. Heat exchanger network synthesis (HENs) can synthesize the optimization of consumption of utility, operating cost, investment cost and network of heat exchangers. With the increasing of the cost of energy, the issues of reducing the consumption of energy become more important nowadays. In the study, we try to use MIHDE to search the optimal solution of sequence HENs approach and simultaneous HENs approach, respectively. It can help to determine the optimal network and minimum total investment cost. Approximate formulas of heat transfer area and constraints are non-linear and non-convexity in the HENs problem. By using conventional mathematics to solve problems, it required transferring primal models to convexity and even giving a moderate initial guess. In the study, we directly apply MIHDE methods to optimize the primal models for global solutions. According to the results of estimations on practical cases using MIHDE and commercial GAMS, we found that the prior is better than the latter for the HEN optimal solution in the study.