基于遗传算法的岩石复电阻率频谱模型参数反演  

Inversion of parameters of rock complex resistivity frequency-spectrum model using genetic algorithm

作　　者：童茂松[1] 

机构地区：[1]中国石油大庆石油管理局测井公司研发中心,黑龙江大庆163412

出　　处：《勘探地球物理进展》2007年第4期271-274,共4页Progress in Exploration Geophysics

基　　金：大庆石油管理局重大科研资助项目"复电阻率测井技术研究"(DQKJZB200318)

摘　　要：复电阻率频谱在地球物理勘探领域具有非常广阔的用途,但是其模型的选择及其参数反演比较困难。遗传算法作为一种全局寻优、简便、容错性强的算法,可以在复电阻率频谱模型参数的反演中发挥重要作用。采用遗传算法,结合各个模型的参数约束,反演了Cole-Cole模型、基于乘积形式的二阶复Cole-Cole模型、基于和形式的二阶复Cole-Cole模型以及Dias模型等4个模型的参数。结果表明,利用遗传算法可以准确求取复电阻率模型参数,对于岩石复电阻率频谱参数的反演及其应用,Dias模型比传统的Cole-Cole模型及基于乘积或基于和形式的二阶复Cole-Cole模型更具有优势,与单纯考虑实部或者虚部的情况相比,同时考虑实部和虚部的反演过程能够取得更好的效果。The frequency spectrum of complex resistivity has shown great potential for use in geophysical prospecting. Selection of model and inversion of its parameters are difficult. Genetic algorithm, which has several advantages such as global optimization, simple, and, tolerance of errors, can be used in the inversion of the complex resistivity model parameters. The genetic algorithm was used to inverse the parameters of Cole-Cole model, multiple Cole-Cole model, summation Cole-Cole model, and Dias model, by taking the restrictions of each model into account. Results show that this method is valuable to inverse parameters of the models. In terms of the number of parameter and its geophysical meanings, Dias model is of potential advantage over the traditional Cole-Cole model, the multiple Cole-Cole model, or the summation Cole-Cole model. Compared to the situation in which only real or imaginary part be used, the inversion of the parameters by combining the real part with the imaginary part yields better results.

关 键 词：岩石 复电阻率 频谱模型 反演 遗传算法 

分 类 号：P631.8[天文地球—地质矿产勘探]