用改进的遗传算法和高斯牛顿法联合反演三维地下水流模型参数

遗传算法在处理非线性优化问题时具有较好的全局搜索性能,但在局部搜索时搜索效率不高,解的精度亦不高,高斯牛顿法在处理非线性优化问题时的性质正好和遗传算法相反,利用遗传算法和高斯牛顿法的优点,用改进的遗传算法和高斯牛顿法联合反演地下水数值模型参数.首先用遗传算法求出地下水模型参数的初值,然后利用这组初值用高斯牛顿法进行数值模型参数的反演,并以一非均质各向同性三维承压非稳定流理想模型为例,结合有限元法讨论了用遗传算法和高斯牛顿法联合反演地下水数值模型参数的过程.计算结果表明,联合参数反演方法,具有收敛速度快、解的精度高的特点,在地下水渗流和水资源评价等领域可广泛应用.

Parameter Identification in a 3-D Groundwater Flow Numerical Model: an Improved Genetic Algorithm and the Gauss-Newton Method

A genetic algorithm (GA) searches in the whole solving space as it deals with nonlinear optimization problems. But in the local solving space, GA is slow and the solution precision is low. The Gauss-Newton Method(GNM) has inverse characters on these points. In this paper, the GA and GNM are used in the parameter identification of ground water flow. GA solves the initial values of parameters. And then, the parameters are identified by GNM. We take 3-dimensional unsteady state flows in an inhomogeneous isotropic confined aquifer as an ideal model, and discuss application of GA and GNM to inverse problem of hydrogeology parameters with finite element method. It is shown that the improved algorithm converges faster and provides higher precision.