CIESC Journal

• 过程系统工程 • 上一篇    下一篇

基于遗传规划的复杂精馏系统综合

王晓红;李玉刚;胡仰栋   

  1. 中国海洋大学化学化工学院,山东 青岛 266003;青岛科技大学化工学院,山东 青岛 266042
  • 出版日期:2006-12-25 发布日期:2006-12-25

Synthesis of complex distillation systems via genetic programming

WANG Xiaohong; LI Yugang;HU Yangdong

  

  • Online:2006-12-25 Published:2006-12-25

摘要: 提出了基于遗传规划的复杂精馏系统的综合方法.针对复杂精馏系统特点并结合化工领域知识,对遗传算子进行了改进,给出了一套独特的编码方法和求解策略.算法利用层次化结构来直观地表达复杂分离流程,同时解决了其他随机算法编码在表达精馏过程时的歧意性问题.在不必给定精馏系统超结构的情况下,能够完成可行域的自动搜寻.实例计算表明,基于遗传规划的综合策略可以用于求解复杂精馏系统的最优综合问题,实现了包括侧线塔、完全热偶塔及同时考虑热集成的各种复杂精馏结构的自动搜索寻优.

Abstract: A genetic programming(GP) based approach was developed for the synthesis of complex distillation systems.Combining with the knowledge of chemical engineering, some evolutionary functors were improved, and a set of special encoding method and solution strategy was put forward according to the characteristics of the complex distillation system.The flow scheme of the complex separation could be expressed directly by using GP’s special hierarchical structure and the problem of different meanings encountered by other stochastic algorithms for expressing distillation system could be solved.Without any superstructures of distillation system, the feasible region could be looked for automatically.The solutions of examples demonstrated that the optimized synthesis problems of complex distillation system could be solved rapidly with the improved algorithm, and the optimal configurations of thermally coupled(linked)side columns,fully thermal coupling column(Petlyuk column)as well as heat integration between the different columns could be searched automatically.