时间域激发极化法最小垂直梯度约束二维反演研究

发布时间：2018-08-13 08:13

【摘要】：时间域激发极化法简称TDIP,它是以不同岩矿石的激电效应的差异为基础,通过观测和研究大地激电效应,以此达到探测地下地质状况的一种分支电法。传统的时间域激发极化法反演通常使用最光滑模型约束的反演,该传统模型约束的算法虽然能稳定收敛,最终能得到稳定解,但是该方法的反演结果对异常体的边界刻画不清晰,反演结果成像分辨率差,对异常体的定位也不精确,进而不利于后期的地质解释。本文研究的时间域激发极化法最小垂直梯度约束反演将对这些问题有所改善,反演结果对异常体更加聚焦。本文正演研究采取有限元方法,基于已有研究的电阻率法正演基础上,通过使用“等效电阻率法”原理,将代码程序改进,可以得到时间域激发极化法正演程序代码,最后,将改进的程序和前人研究的时间域激发极化法三维有限差分正演分别对设计模型进行正演试算,对比正演结果,验证程序的正确性。本文反演基于Zhdanov的聚焦反演方法和刘小军提出了最小垂直梯度泛函的二维MT数据反演方法研究,引入最小垂直梯度泛函,在数据空间反演和最小梯度泛函的基础上,将最小垂直梯度约束泛函代入TDIP二维反演目标函数中,重新推导公式,改写代码,实现本文研究目的。最后,本文通过设计建立电阻率和极化率结构一致模型和电阻率和极化率不一致模型,分别对不同的理论模型正演数据进行反演试算,将计算反演结果与传统反演方法结果进行比较,以此验证本算法能得到稳定收敛的反演结果,对异常体位置的定位更加准确,异常体的边界刻画更为清晰,反演结果成像具有更高的分辨率,聚焦效果更加明显,证明了本反演算法的可靠性和正确性。
[Abstract]:The time-domain induced polarization method is called TDIP. it is a branch electric method based on the difference of the induced electric effect of different rock and ore and by observing and studying the geoelectric effect in order to detect the underground geological condition. The traditional time-domain induced polarization inversion usually uses the most smooth model constraint inversion. Although the traditional model constrained algorithm can converge stably, it can finally obtain a stable solution. However, the inversion results of this method are not clear to the boundary of abnormal bodies, the imaging resolution of inversion results is poor, and the location of abnormal bodies is not accurate, which is not conducive to the later geological interpretation. The time domain induced polarization method with minimum vertical gradient constraints will improve these problems, and the inversion results will focus more on the anomalous bodies. In this paper, finite element method is adopted in the forward modeling. Based on the resistivity method which has been studied, by using the principle of equivalent resistivity method, the code can be obtained by using the principle of "equivalent resistivity method". Finally, the time domain induced polarization method can be used to generate the forward program code. The improved program and the time domain excited polarization method are used to test the design model respectively, and the results of forward modeling are compared to verify the correctness of the program. In this paper, the focused inversion method based on Zhdanov and Liu Xiaojun's research on 2-D MT data inversion of minimum vertical gradient functional are proposed. The minimum vertical gradient functional is introduced on the basis of data space inversion and minimum gradient functional. The minimum vertical gradient constraint functional is added to the objective function of TDIP 2-D inversion, the formula is rededuced, the code is rewritten, and the purpose of this paper is realized. Finally, by designing the consistent model of resistivity and polarizability structure and the model of inconsistency between resistivity and polarizability, the forward modeling data of different theoretical models are retrieved. By comparing the calculated inversion results with those of the traditional inversion methods, it is verified that the proposed algorithm can obtain stable convergent inversion results, and the location of the abnormal bodies is more accurate, and the boundary of the abnormal bodies is more clearly delineated. The inversion result has higher resolution and better focus effect, which proves the reliability and correctness of the inversion algorithm.
【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】：2017
【分类号】：P631.324

【相似文献】