With the development of large gradient wells and horizontal wells, resistivity anisotropy has become an important concern. Most of the inversion problems encountered in geophysical applications are time-consuming. In this letter, we propose a novel approach based on deep learning (DL) to solve the time-consuming problem of the inversion of induction logging data in layered anisotropic formation. A deep neural network (DNN) architecture combined with an adaptive moment estimation (Adam) optimization algorithm is utilized. Batch normalization is used to accelerate the training convergence. Data sets are obtained from forward modeling in different anisotropic formations, by which the DNN is trained, validated, and tested. A series of experiments are carried out to test the accuracy, robustness, and speed of the inversion based on DL in the anisotropic formation. The experimental results indicate that the approach based on DL can achieve the inversion with high accuracy. Furthermore, the trained DNN model for the inversion is two orders of magnitude faster than the most commonly used nonlinear iterative algorithm that achieves the same inversion precision. The trained DNN model considerably reduces the inversion computation time without compromising accuracy. Moreover, the approach overcomes the defect of the dependence of the nonlinear iterative algorithm on the initial values. Our study has shown the promising potential of the approach based on DL for inverting the logging data in anisotropic formations that are difficult to quickly invert by the conventional method.

The traditional design method for terahertz metasurface biosensors is cumbersome and time-consuming, requires expertise, and often leads to significant discrepancies between expected and actual values. This paper presents a novel approach for the fast, efficient, and convenient inverse design of THz metasurface sensors, leveraging convolutional neural network techniques based on deep learning. During the model training process, the magnitude data of the scattering parameters collected from the numerical simulation of the THz metasurface served as features, paired with corresponding surface structure matrices as labels to form the training dataset. During the validation process, the thoroughly trained model precisely predicted the expected surface structure matrix of a THz metasurface. The results demonstrate that the proposed algorithm realizes time-saving, high-efficiency, and high-precision inversion methods without complicated data preprocessing and additional optimization algorithms. Therefore, deep learning algorithms offer a novel approach for swiftly designing and optimizing THz metasurface sensors in biomedical detection, bypassing the complex and specialized design process of electromagnetic devices, and promising extensive prospects for their application in the biomedical field.

Logging while drilling (LWD) plays a crucial role in geo-steering, which can determine the formation boundary and resistivity in real time. In this study, an efficient inversion, which can accurately invert formation information in real time on the basis of fast-forward modeling, is presented. In forward modeling, the Gauss–Legendre quadrature combined with the continued fraction method is used to calculate the response of the LWD instrument in a layered formation. In inversion modeling, the Levenberg–Marquardt (LM) algorithm, combined with the line search method of the Armijo criterion, are used to minimize the cost function, and a constraint algorithm is added to ensure the stability of the inversion. A positive and negative sign is added to the distance parameter to determine whether the LWD instrument is located above or below the formation boundary. We have carried out a series of experiments to verify the accuracy of the inversion. The experimental results suggest that the forward algorithm can make the infinite integral of the Bessel function rapidly converge, and accurately obtain the response of the LWD instrument in a layered formation. The inversion can accurately determine the formation resistivity and boundary in real time. This is significant for geological exploration.

As oil and gas reservoirs progress into the mid-to-late stages of development, produced fluids increasingly exhibit high water-cut and complex flow regimes. Conventional water-cut measurement techniques based on capacitance, conductance, and resistance often face challenges in terms of accuracy, stability, and adaptability. In this study, a novel non-contact broadband microwave system, based on a ridged-horn antenna microwave transmission sensor (RHAMTS), is proposed to achieve highly sensitive full-range (0–100%) water-cut monitoring. The RHAMTS consists of two identical ridged-horn antennas, whose geometries are optimized through analytical design calculations and full-wave finite-element simulations. Numerical simulations are first performed to elucidate the sensing mechanism. Subsequently, static and dynamic experiments are conducted under two representative conditions: emulsified oil-water mixtures and stratified oil-water layers. The results indicate that the broadband spectral signatures of the RHAMTS can effectively characterize water-cut in both emulsified mixtures and stratified oil-water layers. For emulsified mixtures, both amplitude attenuation and phase shift vary systematically with water-cut, and the RHAMTS can still effectively characterize water-cut under saline conditions. For stratified oil-water flow, results from both static and dynamic experiments demonstrate that amplitude attenuation provides more robust features for practical water-cut discrimination. Compared with conventional methods, the proposed RHAMTS offers non-contact operation, rich spectral information, and compatibility with various flow regimes, providing a feasible and efficient approach for water-cut monitoring under complex field conditions.

In this article, the intrinsic propagation mechanisms and dispersion properties of the dominant modes supported by different 2-D graphene-based transmission lines (GTLs) at terahertz (THz) are studied in detail. By employing a complete elliptic integral of the first kind, analytical expressions for the wavenumber of plasmons in different GTLs are obtained from the strongly coupled TM mode surface plasmons that propagate on each individual graphene stripline. Besides, per unit length equivalent circuits of different graphene coplanar transmission lines (GCTLs) are introduced. Again, analytical relations between their effective TM mode equivalent circuit lumped elements and graphene surface conductivity are investigated. Moreover, the effects of different GCTLs versus different parameters of the GCTL models for the wavenumber and circuit lumped elements are analyzed. The proposed analytical expressions could reveal the important mechanisms involved in spatially dispersive plasmon propagation connecting with the intrinsic properties of graphene. The features of GCTLs are detailed, showing their great potential for THz applications in future graphene-based nanoelectromagnetic and nanophotonic devices.

A technique of continuous shaping current waveform to suppress relaxation oscillations (ROs) of distributed feedback (DFB) laser for a high-performance optic system is demonstrated. To effectively suppress ROs, expressions for the shaping current waveform are theoretically derived based on the rate equations and different polynomials for the $3^{rd}$ , $5^{th}$ , and $8^{th}$ order Fourier basis functions are introduced. The convolutional neural network (CNN) is employed to predict the multi-parameter values that determine the results of the shaping input current, which exempt from the difficult and time-consuming process of parameter selection. Prior to training, preprocessing of the data obtained from DFB laser forward simulation using min-max normalization aims to improve the training efficiency of the CNN. The shaping current signals obtained from the CNN predicted parameters are put into the equivalent circuit model for the DFB laser to verify the effectiveness of the shaping current technique and CNN parameter optimization. Afterwards, the shaping current waveform is verified in a time division multiplex passive optical network (TDM-PON) utilizing the DFB laser model as a directly modulated source achieving remarkable performance with low cost. The results show that the high-order continuous shaping current modulated technique can successfully suppress the ROs and enhance the performance of the optic system.