摘要
为了获得详细直观的视电阻率测深数据的反演结果,本文研究了利用视电阻率测深数据的偏导数变化和统计学参数,设计初始模型的反演方法。该方法先判断地层电性结构类型,再圈定目标体位置。利用包含地下地层参数的偏导数信息和一维反演结果的统计学特征,设计出有层参数和地层结构的初始模型。在此基础上使用遗传算法修改模型,配合二维有限元法正演达到拟合误差最小,从而完成电阻率测深数据的二维反演。此法的优点在于初始模型与真实的地层结构有很大的相似性;不需要解大型线性方程组,降低计算量;反演过程中独立计算总体拟合误差和异常体影响区域拟合误差,指导优先变异方向。反演后给出直观的解释结果。对理论模型及实测数据试算证明该方法可以获得较实用的反演效果。
In order to obtain details of inversion results,a new inversion method based on partial derivative and statistics for VES is proposed.The inversion would be carried out in two steps.The first step is to judge the electrical structure of a stratum by statistics and partial derivative,and to define the target location.An initial model of the stratum structure is built on partial derivative of the stratum parameters and statistics of 1D inversion.The second step is to modify the parameters by genetic algorithm and finite element stimulation to decrease fit errors.The advantage of this method is that the initial model for inversion is similar to the actual target,and there is no need to calculate large linear equations,so the calculation amount is much lower.The total fit error and the local fit error caused by anomalies are separately calculated during the inversion to direct the variation.The inversion result is visual.Theoretical model and real data experimental tests demonstrate the effectiveness of the method.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
2012年第6期1006-1013,1024+844,共8页
Oil Geophysical Prospecting
基金
深部探测技术与实验研究专项SinoProbe-09-02--地面电磁探测系统(SEP)研制项目(公益性行业科研专项编号:201011079)
中国科学院工程地质力学重点实验室开放基金项目(KLEG201103)
国家自然科学青年科学基金(41104045)联合资助
关键词
电阻率测深
二维反演
偏导数
统计学
遗传算法
resistivity sounding,2D inversion,partial derivative,statistics,genetic algorithm