The transient electromagnetic (TEM) method is a commonly used, nonintrusive, geophysical method, but inherent mutual induction between the transmitter (TX) and receiver (RX) coils strongly influences the measurements. We have developed an opposing-coils configuration to greatly reduce this effect. Three coils are used in this system. The upper opposing coil is physically the same as the lower TX coil, and they are concentric and parallel to the middle RX coil. A pair of currents with equal amplitudes but reverse directions is injected into the opposing and TX coils. Theoretical calculations in free space show that the received magnetic field by the RX coil is zero, which indicates that the mutual induction effect could be largely reduced. Physical experiments prove that an almost-pure secondary field could be acquired using this system. We have studied an optimal separation between the TX and opposing coils to guarantee that the primary magnetic field is powerful and the instrument is compacted for field work. Then, the efficient exploration depth of this system for typical geoelectric models was simulated to be approximately 15–50 m. Comparisons of simulated responses over highly conductive thick plates in free space and a field test over a culvert structure between this system and EM-47 showed that the system has enhanced sensitivity and lateral resolution. This system can be used in near-surface investigations, e.g., groundwater, environmental, and engineering investigations.
11 August 2023 in IEEE Transactions on Geoscience and Remote Sensing
Despite exhibiting excellent inversion results for synthetic data in Magnetotelluric (MT) inversion, applying deep learning (DL) to directly inverting MT field data remains challenging. In this study, different from most previous works that mainly focus on generating massive representative resistivity models to cover the solutions of the field data or constructing a strong network by employing advanced DL techniques, we provide a new perspective in that a multi-window Savitzky-Golay (MWSG) filter is proposed to first smooth the apparent resistivity and phase derived from the MT field measurements before network prediction. This smoothing operation aims to promote the actual apparent resistivity and phase to be close in morphology and smoothness to the training input data, i.e. to adapt the field data to the training sample data. Then, the smoothed apparent resistivity and phase, instead of the original ones, are fed into the well-trained network for instantaneous inversion. Because we create a set of layered resistivity models with gradual-changing resistivity to act as desired output during network training, it together with the proposed MWSG filter enables this work to achieve smooth inversion. Besides, we introduce Swin Transformer (SwinT) to improve the efficiency of MT DL inversion, based on which a physics-informed neural network (PISwinT) is implemented to enhance the generalization capability. We demonstrate the proposed PISwinT-MWSG smooth inversion method in both synthetic and field MT cases, and it is expected to improve the adaptability and practicability of the DL method to directly solve the inverse problems in MT surveys.
1 September 2019 in Journal of Central South University
Transient electromagnetic method (TEM) is a commonly‐used geophysical method of measuring the secondary electromagnetic field induced by transient pulse sources. Because of the inherent mutual‐induction between TX and RX coils, the measured field always contains the mutual‐induction field, which becomes stronger when TX coil becomes smaller and closer to RX coil. It influences the near‐surface TEM systems, which utilize small TX coils to effectively resolve the near surface explorations. The mixture of the mutual‐induction field with the secondary field at early period leads to a blind region in the shallow surface in TEM exploration. To solve this problem, we applied two concentric and parallel current coils in the TEM system. They are opposing coils, which means that the two coils are physically same, but the currents in them are of equal value and in reverse directions. The RX coil is equidistant to the two opposing coils. By means of this special arrangement, the mutual‐induction field received at the RX coil is zero, which means that the mutual induction effect is efficiently eliminated, and pure secondary field of the conductive underground will be measured. It is proved by theory and model calculation that opposing coils transient electromagnetic method has advantages in lateral resolution. It is quite suitable for small coil TEM system aiming at shallow subsurface detection.
11 September 2024 in IEEE Transactions on Geoscience and Remote Sensing
An ongoing trend seeking to incorporate forward modeling, which involves the physical laws of wave propagation, into the network architecture to improve the generalization capability of the deep learning (DL) inversion method has showcased promising applications. However, directly embedding the time-consuming two-dimensional (2-D) magnetotelluric (MT) forward modeling solved by conventional numerical algorithms to facilitate physics-guided DL MT inversion, which usually necessitates millions of forward operations during a complete training session, is challenging. Hence, in this work, we develop a physics-guided DL inversion method (PGWNet) by constructing a W-shaped DL model and performing a multitask learning strategy. The DL model consists of one encoder and two decoders, where the two decoders are independent of each other and share the encoder. During the training process, two decoders are firstly optimized independently by minimizing the model misfit, quantifying the discrepancy between the predicted and labeled resistivity models, and the data misfit, quantifying the discrepancy between the predicted and labeled MT responses, respectively. When model and data misfits back-propagate to the encoder, they are combined to jointly optimize the encoder. Moreover, to ensure practical application effect, this work builds a set of random synthetic resistivity models with gradually varying resistivity values to delineate realistic subsurface structures. We substantiate the developed PGWNet inversion method using synthetic and actual MT data and benchmark it against a fully data-driven DL inversion method and the conventional least-squares gegularization inversion method. It is anticipated to promote the practicability and applicability of the DL inversion method in practical MT prospecting scenarios.
4 September 2025 in IEEE Transactions on Industrial Electronics
The frequency-domain electromagnetic (FDEM) method leverages the skin depth effect to extract subsurface information at varying depths. However, inductive suppression in the transmitter system reduces transmission current at high frequencies, severely degrading the signal-to-noise ratio (SNR) and exploration performance. Traditional impedance matching methods are insufficient due to limited tunability or lack of continuous adjustment. To address these limitations, this article proposes a switch-controlled capacitor impedance matching network (SCCIMN) for adaptive, continuous impedance matching across 1–100 kHz. First, the circuit structure of the SCCIMN is introduced, along with the equivalent principle of SCC. Then, a detailed parameter design scheme is provided to ensure the SCCIMN meets the specified requirements. Next, an adaptive control strategy is proposed, enabling the SCCIMN to achieve continuous automatic adjustment within the designated frequency range. Finally, simulations and experiments validate the proposed approach, demonstrating that SCCIMN effectively enhances the transmission current, with a current enhancement rate exceeding 100% across the entire frequency band, more than doubling the transmission current. Furthermore, the SCCIMN exhibits excellent dynamic performance: Proportional-integral-derivative (PID)-based frequency switching is completed within 50 ms, while a look-up table (LUT) approach enables near-instantaneous switching and stabilization within a single cycle. Comparative experiments with switched-capacitor array impedance matching networks (SCAIMNs) also confirm that the SCCIMN significantly enhances the SNR. Therefore, the proposed SCCIMN holds considerable potential for practical applications of the FDEM method.