Data depth change characteristic self-adaptive two-dimensional resistivity inversion method and data depth change characteristic self-adaptive two-dimensional resistivity inversion system

Abstract

The invention provides a data depth change characteristic self-adaptive two-dimensional resistivity inversion method and system. The method comprises the steps of constructing data sets of apparent resistivity-resistivity model data pairs of different geoelectric models; constructing an adaptive convolution network with adaptively variable convolution kernel amplitudes and offsets of different layer depths according to resistivity depth change characteristics; constructing an inversion loss function carrying vertical weights of resistivity data items, training an adaptive convolution network controlled by the inversion loss function by using the data set, and establishing a mapping relationship between apparent resistivity data and a resistivity model; inputting observation apparent resistivity data into the trained adaptive convolutional network, outputting a resistivity imaging result through the established mapping relation, achieving earth surface two-dimensional resistivity deep learning inversion, and effectively improving inversion quality, especially inversion precision of a deep area.

Description

technical field [0001] The disclosure belongs to the technical field of two-dimensional resistivity inversion, and relates to a two-dimensional resistivity inversion method and system for adaptive data depth change characteristics. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] Surface two-dimensional resistivity detection is a common geophysical prospecting method. Resistivity inversion imaging is the process of recalculating the resistivity distribution of underground media from observed apparent resistivity data, and is the core issue of resistivity detection. Resistivity inversion is a typical nonlinear problem. At present, the general and mature method converts it into a linear problem by omitting the high-order terms of the objective function, which is easy to fall into local optimum, strong dependence on the initial model, and in...

AI Technical Summary

Problems solved by technology

The depth change characteristics of apparent resistivity data directly lead to: ① It is difficult to effectively capture and distinguish abnormal features, and the network output is blur

Images