CIESC Journal

• 过程系统工程 • 上一篇    下一篇

应用广义简约梯度算法增强遗传算法求解化工冶金相平衡

张会刚;朱庆山   

  1. 中国科学院过程工程研究所多相反应实验室,北京 100080

  • 出版日期:2005-06-25 发布日期:2005-06-25

Calculation of chemical and metallurgical phase equilibria using genetic algorithms reinforced by generalized reduced gradient algorithms

ZHANG Huigang;ZHU Qingshan

  

  • Online:2005-06-25 Published:2005-06-25

摘要: 研究可靠的相平衡计算对化工冶金过程有很重要意义,本文采用广义简约梯度法(GRG)来增强遗传算法(GA)用于相平衡计算过程,在遗传算法演化快结束时引入GRG算子,演化结束后使用GRG精修结果,克服了GA局部搜索能力不强的缺点,加快了搜索速度.KCl-FeCl2体系相平衡计算结果显示这种混合算法(Hybrid Algorithms)能够提高GA效率的同时保留了全局收敛的优点,因此其在化工冶金相平衡计算中将有广泛的应用前景.

Abstract: Since reliable calculations of phase equilibria are very important for chemical and metallurgical processes, studies on algorithms of phase equilibria are of value and interest. In this paper, the generalized reduced gradient (GRG) methods were used to enhance the local search ability of conventional genetic algorithms (GAs). GRG, as an operator, was first introduced to speed up the convergence and at last was used to refine the results of GAs.Because of improved efficiency, higher reliability and flexibility, the hybrid algorithms might have broad application in the calculation of chemical and metallurgical phase equilibria.