用水网络的自动合成方法

将遗传规划(GeneticProgramming,GP)应用于用水网络的合成中,提出了用水网络的自动合成方法。该方法同时优化网络结构和操作变量,获得年费用(包括操作费用和设备投资费用)最小的最优流程。研究了采用遗传规划表达用水网络的编码方法和策略,选择一些通用的单元操作(如塔、分割器、混合器、再生器等)组成遗传规划函数集。该方法的主要优点是能够在可行域内自动寻找较优的网络,而无须预先给定超结构。

An Automatic Approach to Design Water Utilization Network

This paper presents an automatic approach for the design of water utilization network (WUN). Water network design was formulated as an optimization problem of network configuration and design variables where both operating and investment cost were optimized simultaneously. Genetic programming (GP) was used to solve the optimization problem. The encode mode to present WUN configuration was studied. Several general unit operations have been selected, such as column, splitter, mixer, regenerator, etc., which constitute the GP function set. The primary advantage of this approach is the automatic search for potential promising alternatives without any pre-defined superstructures.