用剩余信息量最小准则实现直流电测深数据的二维自动反演

自Pelton等发表第一篇直接电测深数据的二维反演文章以来，电测深数据的二维反演解释已广泛应用于资源勘探和环境监测，并取得一定效果。然而，由于该方法中加入了人为确定的平滑约束，其反演解释结果具有很大不确定性，使其在实际生产中普及应用受到限制。因此，如何在现有方法的基础上进一步改进完善成为该方法普及应用的关键。为了使反演解释结果既忠于实测数据，又合理地反映地下介质的变化，作者在本文中讨论了基于最大熵谱原理的剩余信息量最小准则实现电测深数据的二维自动反演。作者在文中还介绍了正演计算和偏导数求取中，提高精度和计算效率的技巧。用三个数值模式算例对本文的自动反演方法进行了验证。数值模拟结果表明，基于剩余信息量电小准则的二维自动反演给出的解释结果基本不受人为因素影响，能反映地下介质电阻率的详细变化。

IMPLEMENTATION OF THE TWO-DIMENSIONAL AUTOMENTIC INVERSION OF THE DIRECT CURRENT VERTICAL ELECTRIC SOUNDING DATA BY USING OF THE PRINCINE OF THE MINIUM RESIDUAL NIFORMATION

Since Pelton et.al.(1978) published the first paper of the two-dimensional inversion of the direct current vertical electric sounding(VES) data, VES measurements have been widely used in resources exploration and environment monitoring, and achieved fairly effects. The inverted interpretation results show, however, some serious non-determinedness, as there are many randomly determined factors in using smooth constraints in the inversion method, which limits its practical application to field measurements. Therefore, to improve the present inversion method becomes the key to the popularization of the inversion method. In order to ensure the faithfulness of the inversion results to the field measurements, and to assure the reasonableness of the resistivity variation of the subsurface formation, the two-dimensional inversion method of the VES data that based on the minimum residual information in accordance of the principle of the maximum entropy has been addressed in this study. The paper also introduces the techniques of improving the accuracy and computational efficiency of the forward modeling and the derivative computation. Three numerical models have been used to verify the forward code and inversion method. The simulation results show that the two-dimensional inversion method of the VES data implemented in this paper can provide the interpretation results which is independent of the artifacts, while has the detail resistivity variations of the subsurface formation.