NPGA-GW在地下水系统多目标优化管理中的应用

在地下水系统管理问题中，涉及到多个相互冲突的目标函数常常被简化为不同形式的单一目标函数来求解，这种通过单一目标函数的优化方法只能给出一个解，由此确定的方案有时会违背决策者的意愿。而通过多目标优化方法可以得到一系列供决策者权衡选择的解集。将地下水流模拟程序MODFLOW 和溶质运移模拟程序MT3DMS 相耦合，采用基于小生境技术的Pareto 遗传算法进行求解，开发了一个用于地下水系统多目标管理的应用程序NPGA-GW。并将该程序应用于一个二维地下水污染修复问题的多目标优化求解，结果表明，该程序能够在较短的时间内得到一系列Pareto 最优解，解的跨度足够决策者进行适当的选择，具有很好的应用前景。

Application of Niched Pareto Genetic Algorithm to Multi-Objective Optimal Design of Groundwater System

Mostly, conflicting subjects are handled as single subject by several methods in groundwater system management problem. However, these techniques are only capable of finding one solution, which may sometimes contravene decision makers’willingness. On the contrary, multi-objective optimization approaches could provide the decision makers with a set of alternate solutions, being selected for a given site. In this paper, a program named NPGA-GW is developed for groundwater system multiobjective management, based on the niched Pareto genetic algorithm, which is coupled with the groundwater flow simulator MODFLOW and the contaminant transport simulator MT3DMS. NPGA-GW is applied to design pump-and-treat system that solves the multi-objective optimization model for a two-dimensional hypothetical test problem. The results show that a set of Pareto optimal solutions could be found in short term, which spanned trade-off curves wide enough for decision makers. To sum up,NPGA-GW has broad prospects in groundwater system management.