应用遗传算法实现精馏分离序列优化综合

Application of Genetic Algorithm in Distillation Separation Sequence Optimization Synthesis

阐述了采用遗传算法求解具有组合爆炸特征的精馏分离序列综合问题。首先在数据结构上将精馏分离序列抽象为二叉树结构编码 ,然后应用图论方法建立起有效的交叉和变异算子 ,最后对 10组分精馏分离序列综合问题实例进行求解。结果表明 ,遗传算法对最优解命中率达 80 % ,平均空间搜索率仅为 0 0 318。因此遗传算法能够成功解算大规模精馏分离序列优化综合问题。

Genetic algorithm is applied to solve distillation separation sequence synthesis problems which are characteristics of combination explosion. At first, the distillation separation sequence can be described as the binary tree structure coding; and then, according to the method of graph theory, the valid crossover and mutation operators are designed. At last, an example of ten-component distillation separation sequence synthesis problems is solved. The conclusion shows that the rate of successful minimizations from genetic algorithm is equal to 80%, and the average ratio of search space is equal to 0.0318. Therefore, genetic algorithm can solve successfully the large-scale distillation separation sequence synthesis problem.