Traditional Inversion Methods Research Articles

Data-driven acoustic impedance inversion with reweighted L1 norm sparsity constraint

Acoustic impedance (AI) inversion is widely used in geophysics and reservoir prediction. But the traditional impedance inversion method cannot fully exploit the sparse characteristics of geological attributes. There are problems with multiplicity and low resolution. To solve this problem, a data-driven acoustic impedance inversion method with reweighted L1 norm constraints (DRL1) is proposed. In the inversion process, the reweighted L1 norm and local cross-correlation analysis are introduced to solve the above problems. The reweighted L1 norm is introduced as a sparse constraint (RL1) to replace the traditional inversion method which is constrained by L1 norm. The RL1 method can describe more sparsity information and improve the resolution of inversion. In addition, the quality of seismic data plays a decisive role in seismic inversion. We add local cross-correlation analysis to the inversion process. We evaluated the rationality of each sampling point in the seismic data by introducing cross-correlation analysis, controlling for their contribution to the inversion, making inversion results more stable and accurate. The inversion objective function is solved by the alternating direction multiplier method (ADMM) algorithm and soft threshold shrinkage algorithm. Finally, we validate the effectiveness of the proposed method through model tests and field data. The results show that our proposed method not only provides a more accurate portrayal of the stratigraphy, but also yields more accurate inversion results.

Learning-Assisted Inversion for Solving Nonlinear Inverse Scattering Problem

Solving inverse scattering problems (ISPs) is challenging because of its intrinsic ill-posedness and the nonlinearity. When dealing with highly nonlinear ISPs, i.e., those scatterers with high contrast and/or electrically large size, the traditional iterative nonlinear inversion methods converge slowly and take lots of computation time, even maybe trapped into local wrong solution. To alleviate the above challenges, a learning-assisted (LA) inversion approach termed as the LA inversion method (LAIM) with advanced generative adversarial network (GAN) in virtue of a new recently established contraction integral equation for inversion (CIE-I) is proposed to achieve a good balance between the computational efficiency and the accuracy of solving highly nonlinear ISPs. The preliminary profiles composed of only small amount of low-frequency components can be got efficiently by the Fourier bases expansion of CIE-I inversion (FBE-CIE-I). The physically exacted information can be taken as the input of the neural network to recover super-resolution image with more high-frequency components. A weighted loss function composed of the adversarial loss, mean absolute percentage error (MAPE), and structural similarity (SSIM) is used under the pix2pix GAN framework. In addition, the self-attention module is used at the end of the generator network to capture the physical distance information between two pixels and enhance the inversion accuracy of the feature scatterers. To further improve the inversion efficiency, the data-driven method (DDM) is used to achieve real-time imaging by cascading U-net and pix2pix GAN, where U-net is used to replace FBE-CIE-I in the LAIM. Compared with other LA inversion, both the synthetic and experimental examples have validated the merits of the proposed LAIM and DDM.

Transfer Learning-Based Coupling of Smoothed Finite Element Method and Physics-Informed Neural Network for Solving Elastoplastic Inverse Problems

In practical engineering applications, there is a high demand for inverting parameters for various materials, and obtaining monitoring data can be costly. Traditional inverse methods often involve tedious computational processes, require significant computational effort, and exhibit slow convergence speeds. The recently proposed Physics-Informed Neural Network (PINN) has shown great potential in solving inverse problems. Therefore, in this paper, we propose a transfer learning-based coupling of the Smoothed Finite Element Method (S-FEM) and PINN methods for the inversion of parameters in elastic-plasticity problems. The aim is to improve the accuracy and efficiency of parameter inversion for different elastic-plastic materials with limited data. High-quality small datasets were synthesized using S-FEM and subsequently combined with PINN for pre-training purposes. The parameters of the pre-trained model were saved and used as the initial state for the PINN model in the inversion of new material parameters. The inversion performance of the coupling of S-FEM and PINN is compared with the coupling of the conventional Finite Element Method (FEM) and PINN on a small data set. Additionally, we compared the efficiency and accuracy of both the transfer learning-based and non-transfer learning-based methods of the coupling of S-FEM and PINN in the inversion of different material parameters. The results show that: (1) our method performs well on small datasets, with an inversion error of essentially less than 2%; (2) our approach outperforms the coupling of conventional FEM and PINN in terms of both computational accuracy and computational efficiency; and (3) our approach is at least twice as efficient as the coupling of S-FEM and PINN without transfer learning, while still maintaining accuracy. Our method is well-suited for the inversion of different material parameters using only small datasets. The use of transfer learning greatly improves computational efficiency, making our method an efficient and accurate solution for reducing computational cost and complexity in practical engineering applications.

A constrained multi-channel hear-through filter design approach using active control formulations

In many daily communication environments, especially during the pandemic era, partitioning structures are placed between people for security or virus concerns, for example, partitions in classrooms, banks, retail stores, etc. . Distortion occurs when sound propagates through these partitions can cause distractions in communication and degrade the service quality. In the current work, a constrained multi-channel active hear-through system is proposed to provide a solution to provide more natural sound communication environments and prevent sound distortions due to portioning structures. The proposed method can preserve both the spectrum and the spatial distribution of the original sound field (the sound field when the partition is absent). This research opens the potential for future development of multi-channel two-way communication systems and improvement in existing augmented reality techniques. Experimental results show that, compared with the traditional direct inverse filter design method, the proposed method can reduce the comb-filtering effect effectively. The constraints applied to the filter design procedure can also prevent instability caused by the acoustic feedback loop.

3-D Crosswell electromagnetic inversion based on IRLS norm sparse optimization algorithms

Due to the complexity of underground reservoir distribution and the non-uniqueness of geophysical inverse problems, there is still a lack of practical and effective cross-well electromagnetic inversion methods. Our goal is to develop an efficient method to reduce the non-uniqueness of the physical property model recovered in the inversion. Based on a defined framework, we propose a 3-D crosswell electromagnetic sparse inversion based on iterative reweighted least squares (IRLS) method. We perform the inversion using a hybrid norm formulation, where the lp-norm (0 ≤ p ≤ 1) sparse optimization problem can be transformed into a series of weighted l2-norm optimization problems. Compared with the traditional norm inversion method, the sparse inversion method can make more effective use of the known physical information and obtain more accurate abnormal body position and shape. Finally, the effectiveness of sparse inversion method is verified by model test and inversion of measured data in mining area.

Integrating Deep Learning and Deterministic Inversion for Enhancing Fault Detection in Electrical Resistivity Surveys

Clays in fault zones have low electrical resistivity, making electrical resistivity tomography (ERT) effective for fault investigations. However, traditional ERT inversion methods struggle to find a unique solution and produce unstable results owing to the ill-posed nature of the problem. To address this, a workflow integrating deep-learning (DL) technology with traditional ERT inversion is proposed. First, a deep-learning model named DL-ERT inversion that maps apparent resistivity data to subsurface resistivity models is developed. To create target-oriented training data, we use approximately 150 field borehole data acquired from various survey areas in South Korea. The DL-ERT inversion algorithm is based on a U-Net structure and includes an additional network called the borehole mixer to incorporate borehole information when available. The DL-ERT inversion model is trained in three stages: base model training, borehole mixer training, and fine-tuning. Results showed that the fine-tuning model provided the highest prediction accuracy for all test datasets. Next, the prediction of the trained model is used as the initial model for the deterministic inversion method to predict the final subsurface model. The efficiency and accuracy of the proposed workflow are demonstrated in fault detection using a field data example compared with traditional deterministic inversion.

Asymptotic linearized inversion for acoustic variable-tilt transversely isotropic media through elliptic perturbation analysis

The linearized inverse scattering problem for acoustic variable-tilt transversely isotropic (TTI) media is studied. In depicting the acoustic anisotropy behavior of the TTI medium with the P-wave normal moveout velocity (NMO) Vn, density ρ, and anisotropic parameters δ and η, a single-scattering procedure of NMO pressure wavefields is developed through elliptical perturbation analysis. By high-frequency approximation, the singly scattered pressure field is naturally related with an acoustic TTI generalized Radon transform (GRT), in which the TTI scattering patterns are explicitly expressed. We asymptotically invert for the multiple material parameters of an acoustic TTI medium via constructing the GRT backprojection operator. The effectiveness of the proposed approach is demonstrated on the layered model, and its superiority over the traditional acoustic isotropic and VTI GRT inversion methods is illustrated on the overthrust model.

Multiple Thermal Parameter Inversion for Concrete Dams Using an Integrated Surrogate Model

An efficient and accurate method for concrete thermal parameter inversion is essential to guarantee the reliable and prompt thermal analysis results of dams. Traditional inversion methods either suffer from low analysis efficiency or are limited in accuracy. Thus, this paper presents a method for multiple thermal parameter inversion based on an integrated surrogate model (ISM) and the Jaya algorithm. This method replaces finite element analysis with an ISM incorporating three machine learning algorithms, Kriging, support vector regression (SVR), and radial basis function (RBF), to describe the mapping relationship between thermal parameters and structure temperature responses. The input datasets for model training and testing are generated by a uniform design approach. Subsequently, a simple and efficient global optimization algorithm, Jaya, is used to identify the thermal parameters by minimizing the error between calculated and monitored temperatures. The effectiveness and practicality of this method are verified by applying monitored data of two strength grades of concrete in a dam. The verification results indicate that the proposed approach can obtain more accurate inversion results than the above individual models. Compared with these models, the inversion errors using ISM are reduced by 8.45%, 3.93% and 20.85%, respectively for C35 concrete, and by 6.53%, 23.82% and 44.43%, respectively for C40 concrete. Additionally, this approach maintains the powerful computational efficiency of surrogate-based optimization, and compared to the methods that directly invert using swarm intelligence algorithms, the analysis efficiency is improved by about 111.7 times.

An advanced post-necking analysis methodology for elasto-plastic material models identification

The tensile test is widely used for identifying the material plastic flow curve: this is a fundamental step in the design and in material behavior analysis. However, during these tests on ductile metals the necking phenomenon occurs, and the stress and strain states become triaxial and non-uniform. In the scientific literature several characterization methods have been proposed to properly address these aspects.The present work proposes an alternative strategy for identifying the plastic flow curve, focusing on temperature and strain-rate insensitive elasto-plastic materials. This is a hybrid methodology whereby a FE-based inverse method is combined with the 1D description of necking proposed by Audoly and Hutchinson. Indeed, the inverse method employs as target function the specimen profile, but to properly define it a relationship between the original and the deformed configuration of the tensile sample is required, and it is provided by this 1D model.The accuracy and the validity of the proposed strategy was verified on benchmark cases with ad-hoc experimental profiles obtained via numerical simulations. Finally, this method was applied to real case studies and the identified plastic flow curve was compared with the one obtained using a traditional FE-based inverse method having as target the force-stroke curve.

Noninductive Control of Permanent Magnet Synchronous Motor Based on Improved Synovial Observer

In the non-inductive control of permanent magnet synchronous motor, the rotor position information of the motor is usually estimated by the synovial film observer. Synovial observer has the advantages of simple control structure and short response time. The switch function of the control process of the synovial observer is usually the sign function. The back electromotive force signal is extracted by the filter, and then the rotor position information is estimated by the inverse tangent method. However, the traditional switching process of sign function contains a large number of high order harmonics, which will cause the problems of phase difference and amplitude attenuation after the signal is filtered by the filter, and the estimation error of arctangent method is large. Based on the above problems, the traditional switching function is improved in this paper. Continuous sigmoid function is used as the switching function, which can effectively solve the problem of excessive vibration during the operation of the motor. In addition, phase-locked loop technology is introduced to replace the traditional inverse tangential method for rotor position estimation, which further improves the estimation accuracy of the system.

Deep learning inversion method of tunnel resistivity synthetic data based on modelling data

AbstractDifferent water‐bearing geological structures in front of the tunnel face are the main cause of tunnel water inrush disasters, affecting tunnel constructionsafety. Due to the narrow tunnel space and the limited data that can be detected, the traditional linear inversion method for detecting them has multiple solutions. In this paper, we establish a database with complex models, including water‐bearing geological structures usually encountered during tunnel construction. Then we build a complex nonlinear relationship mapping between the tunnel face observation data and the resistivity model through the deep neural network algorithm (electrical resistivity inversion network tunnel, ERTInvNet‐T for short). ERTInvNet‐T first uses a fully connected network to extract the tunnel depth features from the observed data, from which a three‐dimensional geoelectric model of the tunnel is then generated by a three‐dimensional deconvolutional network. At the same time, there is a problem with the uneven spatial distribution of the sensitivity of the data to the model. Therefore, depth weighting information constraint based on the distance factor is added to the loss function, which improves the network algorithm's learning ability for different detection positions of the tunnel. The validity of the proposed method is verified by a large number of numerical simulation experiments.

A hybrid FCN-BiGRU with transfer learning for low-velocity impact identification on aircraft structure

In the work process of an aircraft, its external structure frequently encounters some low-velocity impact events, resulting in some barely visible impact damages inside the structure. Therefore, impact localization and reconstruction are critical for the structure’s health monitoring and reliability analysis. However, traditional inversion methods usually perform poorly in identifying random impact loads due to a series of problems. Inspired by the superior model properties of deep learning algorithms, a novel feature learning-based technique for impact load reconstruction and localization is presented. The proposed method consists of two parts, the first of which is utilized to localize the impact loads and is referred to as one-dimensional full convolution network. The other part aims at reconstructing the impact load and is called attention mechanism-full convolutional network-bidirectional gating recurrent unit-multilayer perceptron. Moreover, a transfer learning technique is introduced to optimize the model structure and process in light of the suggested innovative network model. The experimental piece of this research was an aircraft cutting segment whose structural surface was equipped with a number of piezoelectric sensors. These sensors received impact response signals with differences in amplitude from hammer strikes at various positions. In the work, the process of locating first and then reconstructing the impact load history was followed. And the influence of implementing the transfer learning mechanism on the performance of the impact load history reconstruction model was investigated as well. Three experiments and repeated cross-validation verified the efficacy of the proposed method. The findings indicated that the proposed technique could accurately and quickly detect impact loads at various locations. Simultaneously, it also allowed a perfect reconstruction of the impact loads of different impact energies.

Inverse identification method for determining stress–strain curves of sheet metals at high temperatures

Because of the strong dynamic softening effect of sheet metals at high temperatures, necking occurs early and continues for longer periods. Therefore, obtaining a flow curve after necking has occurred is necessary. The high-temperature flow behavior of materials typically exhibits the coupled effects of work hardening, softening, strain-rate sensitivity, and damage, and using the traditional inverse identification method is quite challenging. This study presents a finite element (FE) inverse identification method that considers work hardening, softening, damage, and strain-rate sensitivity. The method uses a coupled viscoplastic-damage constitutive model as the modified object, and the parameters of the damage equation are iteratively modified by the FE method until the simulated force–displacement curve is consistent with that of the experiment. The identification procedure is demonstrated using 7075 aluminum alloy, and a complete high-temperature flow curve prior to fracture is obtained. The obtained flow curve is applied to analyze the strain field distribution in a hot tensile process and to predict the fracture occurrence and thickness distribution under an isothermal bulging test. Experimental and simulation results are shown to be in good agreement, thus proving the validity of the proposed method.

Combining machine learning and physics modelling to determine the natural cave property with fracturing curves

Currently, it is still challenging to determine the geological parameters of natural caves in deep reservoirs accurately. By inheriting the advantages of Machine Learning (ML) method and physics modelling, a novel ML-Physics method is developed to determine the geological parameters of natural caves based on the data mining of fracturing curves obtained during Hydraulic Fracturing (HF) operation. The computational time of ML-Physics method is divided into two stages, preparation-stage and operation-stage. The preparation-stage happens before HF operation, therefore there is no limitation to the computational time. During this preparation-stage, the implicit relationship between cave property and fracturing curve is generated using ML, which usually fails to ensure the accuracy under different geological and operational conditions. The operation-stage happens during HF operation, in which the computational time is limited because the geological parameters of natural caves are required to be determined in real time. During this operation-stage, the physical modelling based inverse analysis method is carried out, in which the initial value is chosen based on the ML results obtained in preparation-stage. Results show that, with the same target error, the iteration step of ML-Physics method (1 iteration) is much less than that of traditional inverse analysis method (5 iterations). After the same iterations, the error of fracturing curve using the traditional inverse analysis method is 0.40%, while the error using ML-Physics method is 0.02%. Meanwhile, the error of permeability using the traditional ML is up to 10.33%, while the error of ML-Physics method is 0.29%. The present ML-Physics method is potentially useful to optimize the HF design based on the data mining of fracturing curves in real time.

Deep Learning Joint Inversion of Electrical Data for Ahead-Prospecting in Tunneling

Water inrush has become one of the bottlenecks restricting tunnel construction. Among various advanced forecasting techniques, the direct current method is more cost-effective and sensitive to water-bearing structures. It has been widely used in exploring water inrush disasters in practical engineering. Although traditional resistivity linear inversion methods are reasonably practical, they usually suffer from volume effects and cannot accurately locate the location and morphology of water-bearing bodies. Therefore, nonlinear techniques such as deep learning have recently become popular to directly approximate the inversion function by learning the mapping of apparent resistivity data to the geoelectric model. This work presents a novel deep learning-based electrical approach that combines resistivity and polarizability to estimate water-bearing location and morphology. Specifically, we design an encoder-decoder network. A shared encoder extracts features from the input data, two encoders output resistivity, and polarizability models, respectively, and fine-tuned collinear regularization for both outputs reduces solutions’ multiplicity. Compared with traditional linear inversion methods and independent parameter inversion, our proposed joint inversion method shows superiority in locating and delineating anomalous bodies.

Wide-area sound speed profile estimation based on a pre-classification scheme for sound speed perturbation modes

IntroductionThe trend of sound speed profile (SSP) inversion is towards wide-area sound speed estimation. However, the traditional inversion method of dividing the latitude and longitude grids has limitations in terms of significantly lower accuracy when samples are lacking. k-means clustering algorithm (K-means) can divide the training class to achieve high accuracy estimation.MethodThis paper proposes a grid-free pre-classification inversion scheme based on empirical orthogonal function (EOF) vectors. The scheme is based on the K-means to classify the samples according to the perturbation mode of the SSP. After classification, the SSP inversion is carried out using the self-organizing map algorithm (SOM). The experimental sea area is selected from the South China Sea, and the inversion results are evaluated using root mean square error (RMSE) as the criterion.ResultThe inversion results show that the inversion error is 2.1 m/s for the pre-classification solution and 2.7 m/s for the solution without pre-classification, a steady improvement of more than 20% in the inversion error. Accuracy is also improved by 2.14 m/s in the depth range where the sound speed perturbance is greatest.DiscussionThis pre-classification scheme has smaller inversion errors and the classification results are reasonable in terms of distribution in time and space. It provides a feasible solution for SSP inversion in sea areas where samples are lacking.

Change the direction: 3D optimal control simulation by directly tracking marker and ground reaction force data.

Optimal control simulations of musculoskeletal models can be used to reconstruct motions measured with optical motion capture to estimate joint and muscle kinematics and kinetics. These simulations are mutually and dynamically consistent, in contrast to traditional inverse methods. Commonly, optimal control simulations are generated by tracking generalized coordinates in combination with ground reaction forces. The generalized coordinates are estimated from marker positions using, for example, inverse kinematics. Hence, inaccuracies in the estimated coordinates are tracked in the simulation. We developed an approach to reconstruct arbitrary motions, such as change of direction motions, using optimal control simulations of 3D full-body musculoskeletal models by directly tracking marker and ground reaction force data. For evaluation, we recorded three trials each of straight running, curved running, and a v-cut for 10 participants. We reconstructed the recordings with marker tracking simulations, coordinate tracking simulations, and inverse kinematics and dynamics. First, we analyzed the convergence of the simulations and found that the wall time increased three to four times when using marker tracking compared to coordinate tracking. Then, we compared the marker trajectories, ground reaction forces, pelvis translations, joint angles, and joint moments between the three reconstruction methods. Root mean squared deviations between measured and estimated marker positions were smallest for inverse kinematics (e.g., 7.6 ± 5.1 mm for v-cut). However, measurement noise and soft tissue artifacts are likely also tracked in inverse kinematics, meaning that this approach does not reflect a gold standard. Marker tracking simulations resulted in slightly higher root mean squared marker deviations (e.g., 9.5 ± 6.2 mm for v-cut) than inverse kinematics. In contrast, coordinate tracking resulted in deviations that were nearly twice as high (e.g., 16.8 ± 10.5 mm for v-cut). Joint angles from coordinate tracking followed the estimated joint angles from inverse kinematics more closely than marker tracking (e.g., root mean squared deviation of 1.4 ± 1.8 deg vs. 3.5 ± 4.0 deg for v-cut). However, we did not have a gold standard measurement of the joint angles, so it is unknown if this larger deviation means the solution is less accurate. In conclusion, we showed that optimal control simulations of change of direction running motions can be created by tracking marker and ground reaction force data. Marker tracking considerably improved marker accuracy compared to coordinate tracking. Therefore, we recommend reconstructing movements by directly tracking marker data in the optimal control simulation when precise marker tracking is required.

Quantifying electron cyclotron power deposition broadening in DIII-D and the potential consequences for the ITER EC system

The injection of electron cyclotron (EC) waves fulfills a number of important tasks in nuclear fusion devices for which detailed knowledge of the spatial power deposition profile is critical. This deposition profile is commonly determined using forward models such as beam or ray tracing. Recent numerical and experimental studies have shown that small-angle scattering of the EC beam as it passes through the turbulent plasma edge can cause significant broadening of the effective deposition profile, leading to considerable underestimation of the deposition width by forward methods. However, traditional inverse methods to determine the deposition profile from measurements overestimate the deposition profile width due to transport broadening. In this work, we implement three novel methods to resolve the EC power deposition profile from measurements that counteract transport broadening by simultaneously resolving transport and power deposition. We validate their assumptions and compare the results from these methods to the traditional break-in-slope method as well as to the TORAY ray-tracing code in a set of DIII-D discharges spanning five different confinement modes. We show that the four different inverse methods, novel and established, paint a consistent picture of deposition broadening. Specifically, we show that the measured power deposition profile is between 1.6 and 3.6 times wider than the TORAY profiles. Moreover, we show the considerable consequences that this level of broadening can have for ITER.

AVO Inversion Based on Closed-Loop Multitask Conditional Wasserstein Generative Adversarial Network

Neural networks are commonly used for post-stack and pre-stack seismic inversion. With sufficient labelled data, the neural network-based seismic inversion results are more accurate than that use traditional seismic inversion methods. However, in the case of insufficient labeled data, the accuracy of neural networks-based seismic inversion results decreases and is even lower than those based on traditional inversion methods. In addition, the seismic inversion results based on neural networks generally suffer from lateral discontinuity. It further reduces the accuracy of the inversion results. To tackle these problems, we propose a pre-stack seismic amplitude variation with offset (AVO) inversion method based on Closed-Loop Multi-task conditional Wasserstein Generative Adversarial Network (CMcWGAN), which is a GAN-based AVO inversion method. CMcWGAN enables simultaneous and accurate inversion of P-wave velocity ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V_p</i> ), S-wave velocity ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">V_s</i> ), and density ( ρ ). Moreover, it uses the low-frequency information of elastic parameters as a conditional input to alleviate the problem of lateral discontinuity in inversion results. Experimental results of simulated data show that the inversion results based on CMcWGAN have higher accuracy than those based on traditional AVO inversion methods. In addition, when the seismic angle gather is noisy, CMcWGAN has better robustness than the traditional methods. CMcWGAN can also obtain reasonable AVO inversion results in field seismic angle gather data.inversion results. Experimental results of simulated data show that the inversion results based on CMcWGAN have higher accuracy than those based on traditional AVO inversion method. In addition, when the seismic angle gather is noisy, CMcWGAN has better robustness than traditional method. CMcWGAN can also get reasonable AVO inversion results in field seismic angle gather data.

Extraction of Multiple Electrical Parameters From IP-Affected Transient Electromagnetic Data Based on LSTM-ResNet

The induced polarization (IP) effect due to a polarizable body distorts transient electromagnetic (TEM) data, thereby potentially triggering sign reversal phenomena in the measured response. The measured horizontal electric field associated with a grounded-wire TEM is more strongly affected by IP effects than the measured vertical field, meaning that data inversion is more problematic for its component. The traditional inversion method, which assumes a frequency independent resistivity, is complex to extract the chargeability. Yet, the chargeability provides critical information, so it is important to extract the chargeability in addition to the resistivity from IP-affected TEM data. Thus, we proposed a data-driven method based on deep learning to recover the resistivity and chargeability of IP-affected horizontal electric fields. This method, named LSTM-ResNet, combines long short-term memory (LSTM) and a residual network (ResNet) to estimate subsurface electrical properties. Synthetic tests showed that LSTM-ResNet is computationally efficient and accurate for inversion problems. Based on the inverse results with data added noise, we found that a well-trained neural network was not sensitive to noise. A case study was performed by applying LSTM-ResNet to field data collected by a grounded-wire TEM survey at the Kalatongke copper-nickel ore deposit. LSTM-ResNet recovered the simultaneous resistivity and chargeability distributions of subsurface structures from the IP-affected horizontal electric TEM field. The results show a high-chargeability and low-resistivity layer, which was consistent with the lithologic profiles based on drilling cores, indicating the accuracy and robustness of the proposed framework for multi-parameter inversion.