论文部分内容阅读
瑞利波反演属于非线性最优化问题,已有的局部线性化方法(如阻尼最小二乘法)常使解估计陷入目标函数的局部极小值之中,且严重地依赖初始模型。另外,目前被广泛应用的一些简单的瑞利波反演方法也存在明显的缺陷。全局优化法—遗传算法大大放松了对初始模型选择的要求,且不易陷入局部最优解中。本文用已有的简单宜行的反演解释方法获取初始模型,从而确定模型参数的搜索范围,再用遗传算法反演得到最终的介质模型,效果非常理想。
Rayleigh wave inversion is a nonlinear optimization problem. Existing local linearization methods (such as damped least square method) often cause the solution estimate to fall into the local minimum of the objective function, and rely heavily on the initial model. In addition, some simple Rayleigh wave inversion methods widely used now also have obvious defects. Global Optimization - Genetic Algorithm greatly relax the requirements of the initial model selection, and not easy to fall into the local optimal solution. In this paper, the original model is obtained by the existing simple and convenient inversion interpretation method to determine the search range of the model parameters, and then the final medium model is obtained by genetic algorithm inversion. The result is very satisfactory.