论文部分内容阅读
地震成像处理中的剩余静校正估值是一种最优化问题,解决这类复杂问题需要采用较好的方法。遗传算法是一种有意义的方法,它大体上以最优化为基础以保证生物的进化。本文研究了在估算剩余静校正时该算法的适用性,给出了三种新要素,以使其能成功地应用于剩余静校正。这三种要素是:(1)群种裂变为子群种,子群种之间的种类受限制;(2)在不同程度上圈定未校正输入叠加值的搜索区;(3)修改最优化函数,以便考虑随CDP变化的构造特征,当所搜索相空间的体积较大时,引入子群种可加强搜索的效果,关于算法侧
Residual static correction estimation in seismic imaging is an optimization problem, and a better method is needed to solve such complex problems. Genetic algorithm is a meaningful method, it is based on the optimization in general to ensure the evolution of biology. In this paper, we study the applicability of this algorithm when estimating the residual statics, and give three new elements so that they can be successfully applied to the residual statics. These three elements are: (1) species fission into subgroup species, species of subgroup species are limited; (2) to delineate, to varying degrees, the search area for unadjusted input superposition values; (3) to modify the optimization Function in order to account for structural features that vary with CDP. Introducing subgroup species can enhance the search performance when the volume of the searched phase space is large. As for the algorithm side