针对地下水有限元数值模拟中区域三角网格剖分复杂难以处理的情况,提出适合其特点的Delaunay三角网格自动剖分方法,并对含有多个参数分区的含水层进行网格剖分,最后利用遗传算法和有限元程序相耦合来反演含水层水文地质参数。结果表明此方法可大大简化地下水数值模拟的前处理工作,并能提高有限元网格剖分的有效性和准确性,从而得到令人满意的数值模拟结果。
Considering the complexity and difficulty of the regional discretization in finite element groundwater modeling, an automatic discretization method based on Delaunay triangulation (DT) is developed for the purpose. A Delaunay triangulation mesh of an aquifer with multiple hydro--geological parameters is obtained by using this automatic discretization method. Then a genetic algorithm (GA) coupled with finite element method (FEM) is applied to estimate hydro--geological parameters of the inverse problem. The application result shows that the approach does greatly not only simplify the pre-- processing of groundwater flow modeling, but also improve the efficiency and accuracy of the finite element mesh, leading to satisfactory groundwater numerical simulation.