PC-based artificial neural network inversion for airborne time-domain electromagnetic data

Ma, Mingyao; Che, Hongwei; Yang, Er-Wei; Ji, Yanju; Yu, Shengbao; Lin, Jun

doi:10.1007/s11770-012-0307-7

Cited by 26 publications

(14 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An ANN is a parallel system consisting of many interconnected units. Zhu et al (2012) use an ANN based on principal components as a proxy for a look-up table to relate the response to the earth parameters and avoid the calculation of Jacobian derivatives.…”

Section: Other 1d Inversion Methodsmentioning

confidence: 99%

“…In comparison with the GaussNewton technique, which solves the normal equation derived from a quadratic approximation to equation 20, a gradient-based optimization algorithm such as NLCG directly minimizes the penalty functional of equation 19 (Rodi and Mackie, 2001;Newman and Boggs, 2004;Avdeev, 2005;Kelbert et al, 2008;Egbert and Kelbert, 2012). For this purpose, Liu and Yin (2013) suggest the following iterative procedure:…”

Section: Nonlinear Conjugate Gradient Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Review on airborne electromagnetic inverse theory and applications

et al. 2015

View full text Add to dashboard Cite

Airborne electromagnetic (AEM) data processing and inversion has progressed from the early conductivity-depth imaging, to 1D inversion for a layered earth model, and most recently to 2D/3D inversions. The new processes have used methods for searching for solutions, applying constraints, and focusing images, such as Occam's, laterally constrained, and holistic inversions. For airborne data imaging and 1D inversion, some algorithms are fast, whereas others have control over data misfit and the vertical or lateral smoothness. Beyond 1D algorithms, 2D/3D inversions are based on currently used methods such as isolated conductor models (which are good for highly conductive orebodies in resistive background) and techniques that discretized regions that are used for more complex structures and backgrounds. In the latter case, the computations are slow, so research is focusing on time-efficient computer algorithms for solving the equations such as Gauss-Newton, quasi-Newton, and nonlinear conjugate gradient algorithms. For the electromagnetic (EM) problem, the solution can be obtained in a more practical time frame if material that has minimal impact is ignored. Two approaches can be used to speedup the EM inversion process -the moving footprint and direct solver methods. We hope our work will to some extent help stimulate and focus the research in AEM inversion.

show abstract

Section: Other 1d Inversion Methodsmentioning

confidence: 99%

Section: Nonlinear Conjugate Gradient Methodsmentioning

confidence: 99%

Review on airborne electromagnetic inverse theory and applications

et al. 2015

View full text Add to dashboard Cite

show abstract

“…Because of the nonstationarity of the geomagnetic field, these methods are not suitable for geomagnetic SNR estimation. Zhu (2012Zhu ( , 2013 and Wang (2015) analysis needs observation data in at least three groups at the same observatory, which is improbable for most geomagnetic observatories. Jiang (2013) used maximum likelihood estimation to calculate geomagnetic noise through multiple iterations.…”

mentioning

confidence: 99%

Background noise estimation of geomagnetic signal

Yao¹,

Zhang²,

Teng³

et al. 2018

Preprint

View full text Add to dashboard Cite

A fast Fourier transform was applied to fit the geomagnetic diurnal variation. Fitting results showed that when the polynomial degree was greater than 160, the residual error was close to 0 nT. White noise is the main component of the residual error when the polynomial degree was greater than 160, so this method was adopted to calculate the background noise of the 10 geomagnetic field. Spectrum analysis further demonstrated the necessity to remove background noise from geomagnetic data.

show abstract

“…The authors discuss 1D AEM inversion using the following methods: least-squares inversion, Occam's inversion, laterally constrained inversion, holistic inversion, Bayesian inversion, and simulated annealing as well as a short discussion of extensions of Zohdy's method, the S-inversion method, and a method by Sattel (2005), and an artificial neural net method of Zhu et al (2012). The diverse range of topics includes a tutorial on EM inversion methods, a new method for relating microseismicity to diffusivity in fluid injection, a new approach to deblending blended seismic data, a joint inversion of seismic and resistivity, a new seismic-dispersion analysis method, a rock-physics study of the effects on geophysical measurements of micrite in carbonates, a study of shear modulus determination in heavy oils, and a new EM transmitter-array design.…”

mentioning

confidence: 99%

“…The authors discuss 1D AEM inversion using the following methods: least-squares inversion, Occam's inversion, laterally constrained inversion, holistic inversion, Bayesian inversion, and simulated annealing as well as a short discussion of extensions of Zohdy's method, the S-inversion method, and a method by Sattel (2005), and an artificial neural net method of Zhu et al (2012). The following imaging methods are discussed: differential resistivity imaging, conductivity-depth imaging, and the EMFlow method, each of which is not an "inversion" in the sense of solving an inverse problem through some iterative technique, under some norm.…”

mentioning

confidence: 99%

Geophysics Bright Spots

2015

The Leading Edge

View full text Add to dashboard Cite

O ptimization, rock physics, and improved EM survey design compose the July-August 2015 Geophysics constellation of "Bright Spots." The diverse range of topics includes a tutorial on EM inversion methods, a new method for relating microseismicity to diffusivity in fluid injection, a new approach to deblending blended seismic data, a joint inversion of seismic and resistivity, a new seismic-dispersion analysis method, a rock-physics study of the effects on geophysical measurements of micrite in carbonates, a study of shear modulus determination in heavy oils, and a new EM transmitter-array design.Yin et al. present a "Review on airborne EM inverse theory and applications" as a tutorial on airborne EM (AEM) imaging and inversion methods. The following imaging methods are discussed: differential resistivity imaging, conductivity-depth imaging, and the EMFlow method, each of which is not an "inversion" in the sense of solving an inverse problem through some iterative technique, under some norm. The authors discuss 1D AEM inversion using the following methods: least-squares inversion, Occam's inversion, laterally constrained inversion, holistic inversion, Bayesian inversion, and simulated annealing as well as a short discussion of extensions of Zohdy's method, the S-inversion method, and a method by Sattel (2005), and an artificial neural net method of Zhu et al. (2012). The authors also discuss 2D and 3D AEM inversion methods assuming an isolated conductor. These include application of the Gauss-Newton method, the nonlinear conjugate gradient method, and quasi-Newton methods. The authors discuss a collection of methods created by the CSIRO group in Australia: the LeroiAir, ArjunAir, SamAir, and LokiAir algorithms. Finally, the authors discuss acceleration techniques aimed at improving computational speed and storage.In "Bayesian inversion of pressure diffusivity from microseismicity," Poliannikov et al. develop a probabilistic model that ties f luid pressure during injection (in hydraulic fracturing) to observed microseismicity. The authors propose a Bayesian reformulation of the (Shapiro et al., 2002;Shapiro et al., 2005aShapiro et al., , 2005b physical pore-pressure/seismicity model, which is based on the Mohr-Coulomb theory of rock fracturing. The authors' approach produces a model that captures the essential parameters that control f luid pressure, rock failure, and seismic wave propagation that can be used for forward predictions and inversions. The authors also present a probabilistic framework that allows rigorous uncertainty analysis, which quantifies prediction and inversion errors, and helps the user to find ways to improve the quality of predictions by identifying dependences between uncertainties at different stages of the model. The inversion is for diffusivity of the fracture system during injection, which the authors stipulate is likely to be inadequate to forecast production capacity, for example.Cheng and Sacchi address the problem of deblending data generated from simultaneous seismic sources in "Se...

show abstract

PC-based artificial neural network inversion for airborne time-domain electromagnetic data

Cited by 26 publications

References 18 publications

Review on airborne electromagnetic inverse theory and applications

Review on airborne electromagnetic inverse theory and applications

Background noise estimation of geomagnetic signal

Geophysics Bright Spots

Contact Info

Product

Resources

About