时间域激发极化法电阻率和极化率二维同步反演研究

发布时间：2018-11-19 07:38

【摘要】：时间域激发极化法(Time-domain induced polarization method,简称为TDIP)是以岩、矿石的激电效应差异为物质基础,通过观测和研究大地激电效应,来探查地下地质情况的一种电法勘探分支方法。该方法在寻找金属矿、能源和工程地质等不同领域均有着广泛的应用,尤其在寻找浸染状矿产资源方面效果显著。国内外学者对TDIP二维反演算法进行了大量研究,但是都是采用分步反演思路,即先反演出电阻率之后,在所反演出的电阻率模型基础之上再反演极化率参数,这样就存在着极化率反演结果严重依赖于电阻率反演结果等问题。为了解决上述问题,本文在Gallardo和Meju所提出交叉梯度理论的基础之上,研究了时间域激发极化法电阻率和极化率二维同步反演的算法。本文的正演计算过程是利用有限单元法和“等效电阻率法”,计算得到视电阻率和视极化率参数。本文在电阻率法反演基础之上完成TDIP的反演算法。Weerachai Siripunvaraporn和Gary Egbert在OCCAM反演的基础上进行了改进,将反演从模型空间域转换到了数据空间域进行,该方法降低了反演过程中方程的阶数,并且不需要求取模型协方差的逆矩阵,从而大大减少了所需的计算时间和计算内存,提高了反演的效率。本文将数据空间反演应用到了二维TDIP反演中,研究了较成熟的电阻率法数据空间域反演算法,并在此基础上实现了数据空间域的TDIP反演算法。在完成了TDIP二维分步反演算法之后,本文实现了基于交叉梯度约束的数据空间域TDIP二维同步反演算法,并利用电阻率极化率物性分布一致和不一致模型的合成数据进行了反演试算。结果表明:对于不同的模型而言,同步反演比分步反演对异常体的空间形态有更好的恢复,其中低阻高极化体的同步反演结果电阻率和极化率均更接近于真实数值,高阻高极化体的同步反演结果不仅在数值上有所改善,其空间分布范围也更加准确,而电阻率和极化率结构不一致的模型反演结果进一步表明了同步反演能改善反演结果。总之,同步反演能改善电阻率和极化率的单独反演结果,这进一步证明了同步反演算法具有更普遍的适用性。
[Abstract]:The time-domain induced polarization method (Time-domain induced polarization method,) is a branch of electrical prospecting method based on the difference of the induced electric effect between rock and ore. The geoelectric effect is observed and studied to explore the underground geological conditions. This method has been widely used in the exploration of metal ore, energy and engineering geology, especially in the exploration of disseminated mineral resources. Scholars at home and abroad have done a lot of research on the two-dimensional inversion algorithm of TDIP, but they all adopt the idea of step inversion, that is, inversion of resistivity and then inversion of polarizability parameters on the basis of the inverse resistivity model. In this way, the results of polarizability inversion depend heavily on the resistivity inversion results and so on. In order to solve the above problems, based on the cross-gradient theory proposed by Gallardo and Meju, the time-domain induced polarization method for 2-D simultaneous inversion of resistivity and polarizability is studied in this paper. In this paper, the forward calculation process is to calculate the apparent resistivity and apparent polarizability parameters by using the finite element method and the "equivalent resistivity method". In this paper, based on resistivity inversion, TDIP's inversion algorithm. Weerachai Siripunvaraporn and Gary Egbert are improved on the basis of OCCAM inversion. The inversion is transformed from model spatial domain to data spatial domain. This method reduces the order of the equation in the inversion process and does not need to obtain the inverse matrix of the covariance of the model, thus greatly reducing the computation time and computing memory required, and improving the efficiency of inversion. In this paper, the data space inversion is applied to two-dimensional TDIP inversion, and the mature inversion algorithm of resistivity method in data space domain is studied. Based on this, the TDIP inversion algorithm in data space domain is implemented. After completing the TDIP 2-D step inversion algorithm, the TDIP 2-D synchronous inversion algorithm in data space domain based on cross-gradient constraint is implemented in this paper. The inversion experiments are carried out by using the synthetic data of consistent and inconsistent physical distribution of resistivity polarizability. The results show that, for different models, the simultaneous inversion of the score step has a better recovery to the spatial shape of the anomalous body, and the resistivity and the polarizability of the low-resistivity high-polarimetric body are closer to the real value. The synchronous inversion results of high resistivity and high polarizability are not only improved numerically, but also the spatial distribution range is more accurate. The results of model inversion with different resistivity and polarizability structure further show that synchronous inversion can improve the inversion results. In a word, synchronous inversion can improve the single inversion results of resistivity and polarizability, which further proves that the synchronous inversion algorithm is more universal.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.3

【相似文献】