基于标准遗传算法的地下水污染源溯源方法

A Groundwater Pollution Source Identification Method Based on the Simple Genetic Algorithm

本文将标准遗传算法应用于确定一维均质含水层中污染源位置,利用数值实验分析污染源预估域对溯源计算时间、均值及方差的影响。研究发现,基于标准遗传算法的溯源方法可以在文中的理想算例中取得较好的溯源效果。溯源时间在整体上呈现溯源问题越复杂溯源时间越长的规律。延污染物运移主方向上,污染溯位置的变化对溯源结果的影响小于垂直于污染物运移主方向上的影响,而溯源方法陷入局部最优的可能要大于在垂直于污染物运移主方向上陷入局部最优的可能。

The simple genetic algorithm is applied to find the pollution source location in groundwater. Numerical test is used to find the influence of estimated pollution range on the time consuming, mean, and standard deviation of identification result. The bigger the estimated range is, the more time is consumed. A slight movement of the estimated source in the direction perpendicular to the major migrate direction leads to big bias between the calculated source location and the real location. The chance that optimization model falls into the local optimum location is growing in the major migration direction.