Long-term health monitoring of unattended sensing nodes is essential for remote Earth Observation Networks (EONs), yet it remains challenging because anomaly-induced response drifts often resemble genuine geophysical variations in both spectral and morphological characteristics. This ambiguity makes sensor degradation difficult to distinguish from valid observations, particularly when measurements are affected by regional spatiotemporal coupling and labeled fault data are unavailable. To address these issues, we propose a physics-informed and data-driven anomaly detection framework for EON sensing nodes. Multi-domain complementary features describing amplitude, spectral, and phase behaviors are constructed to improve the separability between natural geophysical variability and sensor-induced distortions. Building on these features, a Memory-Enhanced Transformer-Graph Convolutional Network (ME-TGCN) is developed to model spatiotemporal dependencies and disentangle node-specific abnormal responses from shared regional variations, while an external memory mechanism preserves long-term healthy operating patterns. Residual modeling errors are further compensated to improve normal-response estimation. Anomalies are then identified from the discrepancy between predicted and observed sensor responses through Mahalanobis-distance-based residual analysis with adaptive thresholding, enabling detection without labeled fault samples. Experiments on real-world EON datasets show that the proposed method outperforms representative baseline methods in both normal-state prediction and anomaly detection, and can reliably track progressive sensor degradation and localized anomalies. Cross-regional transfer results further demonstrate its robustness, generalization capability, and practical value for large-scale unattended sensor network monitoring.

With the rapid increase in number of electric vehicle (EV) charging facilities, traditional on-site verification methods for the actual loads face many challenges, particularly in large-scale measurements of near-retirement charging facilities or in extreme environments. Important factors are the measurement efficiency, measurement performance prediction, and a shift from periodic verification to inaccurate replacement. In response to this challenge, this article developed a time series model based on series encoding and dual-pooling feature construction. The model encodes time series information into image information and extracts discriminative features by designing a residual module based on the dual pooling features and visual pattern capture module. Then, the proposed prediction model was integrated into the in-situ measurement system to predict the energy measurement errors and health status of EV charging facilities, implement flexible calibration cycles, and ultimately reduce labor and material costs. Finally, combined with the Monte Carlo Dropout approximate Bayesian process, the uncertainty of the prediction results was evaluated. Validated on a Python/PyTorch-based platform, the proposed model achieves an mean squared error (MSE) of 0.0068 in electric energy error prediction, outperforming five advanced deep neural networks by 4.2253%-67.6190%. It exhibits high training efficiency (99.97% faster than iTransformer) and strong robustness (76.1% lower mean absolute error under Low state-of- charging). Ablation studies confirm that the self-attention, stacked autoencoder, and dual-pooling attention modules improve MSE by 13.9241%, 8.1081%, and 12.8205%, respectively. Moreover, the prediction uncertainty of 0.298% meets metrological standards for EV charging facilities.

Shapley value is a widely used tool in explainable artificial intelligence (XAI), as it provides a principled way to attribute contributions of input features to model outputs. However, estimation of Shapley value requires capturing conditional dependencies among all feature combinations, which poses significant challenges in complex data environments. In this article, EmSHAP (Energy-based model for Shapley value estimation), an accurate Shapley value estimation method, is proposed to estimate the expectation of Shapley contribution function under the arbitrary subset of features given the rest. By utilizing the ability of energy-based model (EBM) to model complex distributions, EmSHAP provides an effective solution for estimating the required conditional probabilities. To further improve estimation accuracy, a GRU (Gated Recurrent Unit)-coupled partition function estimation method is introduced. The GRU network captures long-term dependencies with a lightweight parameterization and maps input features into a latent space to mitigate the influence of feature ordering. Additionally, a dynamic masking mechanism is incorporated to further enhance the robustness and accuracy by progressively increasing the masking rate. Theoretical analysis on the error bound as well as application to four case studies verified the higher accuracy and better scalability of EmSHAP in contrast to competitive methods.

To address the high cost and low efficiency of electric vehicle (EV) charging facility verification, this study proposes a physics-informed and data-driven method for concomitant in situ error estimation and uncertainty quantification. First, based on the operational mechanism and topology of concomitant in situ measurement, an adaptive systematic error compensation model is constructed using temperature and charging power. Then, using multisource historical data generated by this mechanism model, a CV-TCB prediction method [(CEEMDAN-VMD-Transformer-CNN-BiLSTM (CV-TCB)] is proposed. It first performs deep decomposition and feature fusion on multisource time-series data (e.g., energy measurement error and temperature) via complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN)-variational mode decomposition (VMD); then integrates Transformer for long-range dependencies, CNN for local features, and bidirectional LSTM (BiLSTM) for bidirectional temporal modeling, improving prediction accuracy and robustness. A Bootstrap-based uncertainty quantification method incorporates data noise and model errors into interval estimation, providing reliable decision support for error assessment. Experimental results show that the proposed CV-TCB model achieves a root-mean-square error (RMSE) of 0.0227 and an R-2 of 0.9760. Compared to state-of-the-art hybrid benchmarks, it reduces RMSE by 10.6% and increases R-2 by 0.62%. Ablation and sensitivity analyses validate the model's effectiveness and robustness. Moreover, the expanded uncertainty is U=0.340 % ( k = 2 ), demonstrating high reliability and stability.

Abstract To address the difficulty of online detection of metering errors in electric vehicle (EV) charging stations, this paper proposes an error identification method that integrates dynamic data estimation with topological structure analysis. The method relies solely on remotely collected data from station-level master meters and pile-level sub-meters. It constructs a generalized error framework that incorporates a dynamic conversion efficiency model and line-loss estimation, and employs a generalized ridge regression algorithm based on RSD grid optimization to mitigate multicollinearity. In addition, an uncertainty quantification mechanism is embedded to provide comprehensive characterization of measurement results and output credible intervals, thereby enabling stable estimation of metering errors. Our findings reveal that the method effectively quantifies and isolates the impacts of various interfering factors on metering deviations, significantly reducing estimation errors for station operators caused by inaccurate metering. Moreover, the method is independent of specific station configurations, demonstrating strong generality. Finally, validated through simulations and real-world scenarios and compared with several typical methods, the results confirm that the proposed method offers marked advantages in estimation accuracy, robustness, and adaptability. Experimental results demonstrate that the proposed method achieves a parameter RMSE of 0.0700, a model R 2 of 0.9898, and an expanded uncertainty ( k = 2) ranging from 0.218% to 0.623%. Across 100 simulation datasets and two real-world charging stations, it reduces parameter estimation error by over 70% compared with baseline methods (LMRLS, DREM-DRLS, NNF-GDRLS, and NNF), and the maximum deviation from on-site calibration errors does not exceed 0.46%, with all validation outcomes fully consistent with the on-site calibration results. The method is independent of specific station configurations, and offers marked advantages in estimation accuracy, robustness, and adaptability. This study provides a new data-driven pathway and a practical solution for online metering performance monitoring of EV charging stations.

High-performance terahertz (THz) sensing remains challenging due to the intrinsically weak interaction between terahertz waves and matter, especially for ultrathin or low-concentration analytes. In this work, we propose and systematically investigate an all-dielectric terahertz metasurface sensor based on dual Fano resonances. The metasurface consists of dissimilar silicon-based resonators integrated on a silicon dioxide substrate, enabling the excitation of two narrowband transmission resonances within the 1.15-1.55 THz frequency range. Comprehensive parameter sweeps combined with transmission maps and electric-field analyses reveal that the resonances originate from strong near-field coupling and mode hybridization between adjacent dielectric elements, leading to enhanced field localization and suppressed radiative losses. The sensing performance is evaluated by monitoring the resonance frequency shifts under varying ambient refractive indices. The higher-frequency resonance exhibits a sensitivity of up to 196.5 GHz RIU-1 and a figure of merit exceeding 217.8, while maintaining good linearity and robustness over a wide refractive-index range. Compared with previously reported Fano-type terahertz sensors, the proposed design achieves a favorable balance between quality factor and sensitivity without relying on stringent electromagnetically induced transparency or bound state in the continuum excitation conditions. In addition, a feasible silicon-based fabrication route is discussed, highlighting the practical potential of the proposed metasurface for high-performance terahertz refractive-index sensing.

A configurable terahertz photonic topological insulator (PTI) composed of six rotatable triangular dielectric rods in the hexagonal lattice is proposed. By adjusting the rotation angle of the triangular rods and their distance from the unit cell center, we find that different types of the PTIs with controllable topological phases in high-frequency bandgap can be obtained. The first type experiences topological phase transition at a position that is not a high symmetry point in the Brillouin zone. It shows non-zero Berry curvature near the transition point but vanishing valley Chern number, accompanied with dispersion flattened topological edge states. The second type presents typical valley Hall phase with small or large valley Chern number under specific configurations and the edge states are extremely stable. Besides, both types support high-Q topological corner states. This work provides possible guidance for the application of terahertz configurable topological waveguides and high-Q resonators.

Graph neural networks (GNNs) have emerged as powerful tools for industrial soft sensing, offering the ability to model complex relationships among process variables. However, existing GNN-based soft sensors suffer from two critical limitations: (i) they lack hierarchical modeling capabilities to reflect the multi-level structure of industrial processes, and (ii) their black-box nature hinders interpretability, making it difficult to evaluate the influence of process variables on key performance indicators. These limitations reduce the reliability of GNNs in safety-critical industrial processes. To address these challenges, this paper proposes a tree structure guided GNN framework that integrates a data-driven adjacency matrix with a tree-based process representation derived from prior knowledge, enabling hierarchical feature aggregation and clearer correlation modeling. Based on this, a tree structure guided graph information bottleneck (GIB) method is developed to extract critical subtrees that preserve predictive information while suppressing task-irrelevant redundancies, thereby enhancing interpretability. Furthermore, an intrinsic counterfactual explanation module is introduced to generate actionable anomaly regulation suggestions by identifying minimal and interpretable process adjustments needed to restore key performance indicators to safe operating ranges. Experimental results on simulation and real-world industrial datasets demonstrate that the proposed framework achieves superior predictive accuracy, interpretability, and practical effectiveness in anomaly regulation.

A hollow triangular rod-type valley-Hall photonic topological insulator is proposed, and two tandem residual deep neural networks are built for multimodal inverse design of the structure. One of them is a tandem multilayer perceptron, and the other is a composite tandem network based on variational auto-encoder. The former is used to inversely infer the value of the structural sizes, and the latter is used to predict the structural image of the lattice from demanded design targets. Residual connections are included in both networks to speed up the training convergence as well as avoid vanishing gradient problem. Based on an arbitrary inversely designed lattice, domain walls between two photonic topological insulators with different topology are constructed, and full-wave simulations on the transmission properties are conducted. Numerical results show that robust topologically protected wave propagation is supported along the domain wall with little backscattering, demonstrating that the proposed methods are valid.

A terahertz band-switchable photonic topological insulator (PTI) composed of a C3-symmetric rod-type photonic crystal is designed. By tuning the size of the central cylinder in the lattice, a topological phase transition can occur in the PTI, and the topological nontrivial bandgap can be switched from the first to the second bandgap. In both cases, before and after switching, topological edge-state transport of terahertz waves along zigzag topological domain walls, as well as terahertz corner-state localization in constructed resonant cavities, are numerically demonstrated. In addition, an existence of the topological phase transition is also confirmed when tuning the central unit in the lattice of another C3-symmetric hole-type photonic crystal. This work provides a new approach for flexible terahertz waveguiding and lasing applications.

The microseismic monitoring system relies on multiple geophones to detect seismic events, while other methods to determine the operation status of geophones mainly rely on manual inspection of each geophone or comparison of significant changes in observation data, resulting in lower efficiency and accuracy. To address these limitations, we propose an innovative online detection method based on geophone spatiotemporal correlation and data augmentation to continuously monitor geophone status. The proposed extracts time-frequency and energy distribution features from observation data by applying a 230-Hz low-pass filter to preserve the main frequency band and decompose multiple frequency bands. To enhance the dataset, we use Monte Carlo method to generate additional samples of energy distribution features and extend the time-frequency features from 96 samples to 1000 samples using a generative adversarial network (GAN) model. In addition, a dual-stream spatiotemporal network model is established for detecting geophone states, which utilizes the spatiotemporal correlation between geophones to improve detection accuracy. The accuracy of the model on the simulated dataset is 98.67%, with an F1-score of 0.9834. Using bootstrap to estimate the performance of the model on real datasets, the average accuracy is 98.99%, with a 95% confidence interval of [0.9688, 1.0000]. The experimental results verify that this method can detect geophone anomalies online, reducing the need for manual intervention. In addition to microseismic monitoring, our method also has potential applications in the detection and maintenance of operational sensors.

Industrial computed tomography (CT) is widely used in the measurement field owing to its advantages such as non-contact and high precision. To obtain accurate size parameters, fitting parameters can be obtained rapidly by processing volume data in the form of point clouds. However, due to factors such as artifacts in the CT reconstruction process, many abnormal interference points exist in the point clouds obtained after segmentation. The classic least squares algorithm is easily affected by these points, resulting in significant deviation of the solution of linear equations from the normal value and poor robustness, while the random sample consensus (RANSAC) approach has insufficient fitting accuracy within a limited timeframe and the number of iterations. To address these shortcomings, we propose a spherical point cloud fitting algorithm based on projection filtering and K-Means clustering (PK-RANSAC), which strategically integrates and enhances these two methods to achieve excellent accuracy and robustness. The proposed method first uses RANSAC for rough parameter estimation, then corrects the deviation of the spherical center coordinates through two-dimensional projection, and finally obtains the spherical center point set by sampling and performing K-Means clustering. The largest cluster is weighted to obtain accurate fitting parameters. We conducted a comparative experiment using a three-dimensional ball-plate standard. The sphere center fitting deviation of PK-RANSAC was 1.91 μm, which is significantly better than RANSAC’s value of 25.41 μm. The experimental results demonstrate that PK-RANSAC has higher accuracy and stronger robustness for fitting geometric parameters.

To improve the long-distance measurement accuracy of the total station, an optimization method for the meteorological correction formula is proposed. Analysis revealed the potential for optimizing this formula. Adjusting the formula coefficients was approached as an optimization problem to minimize residual errors. Environmental parameter sensors recorded real-time data along a long baseline during continuous ranging. This data was then used to construct optimization functions. Various combinations of formula coefficients were optimized and validated. The mean ranging error was reduced by 99.8 %, and the standard deviation of the error decreased by 81.2 %. To verify the method, the optimized coefficients were applied to other measurement periods, resulting in a residual error reduction of over 37 %. The study demonstrates that optimizing the meteorological correction formula coefficients is a highly effective and promising method for significantly reducing field ranging errors.

The PT symmetric phase transition existing in the gain-loss boundary of non-Hermitian second-order topological photonic crystals is investigated. The variation of the 2D bulk polarization of the photonic bands is analyzed as the gain-loss quantity increases. By tuning the gain or loss strength, topological edge states can appear at the interface between different non-Hermitian topologically nontrivial photonic crystals. Both the edge and corner states are studied in a topological cavity composed of a square closed domain wall between gain and loss domains. Calculated results show the non-Hermitian cavity has better mode localization of the electric field than the Hermitian one. Besides, the existence of the non-Hermitian induced states does not require strict balance of the gain and loss.

Predicting surface-wave travel-time shifts is valuable for analyzing potential effects caused by changes in medium properties, station clock errors, instrument response errors, and other factors. Many current neural networks used in seismology are single-station models trained using single-station (pair) data. However, most seismic methods require knowledge of the spatial positions between multiple stations. Multiple stations contain rich interrelationships and spatial information that cannot be exploited by single-station models. We proposed a multistation neural network structure Transformer Graph Convolutional Network (TGCN) that utilizes temporal attention and spatial attention to capture spatiotemporal information for predicting relative travel-time shifts. Before that, we introduced a method that treats station pairs as nodes and constructs a graph with multiple station pairs. We collected original ambient noise waveforms from 2017 to 2019 in the Alaska region and 2010 to 2014 in the southern California region to obtain relative travel-time shift sequences of station pairs for model training and testing. To showcase the improvement of spatial information to the model, we compared TGCN with two other baseline single-station models—temporal convolutional network and long short-term memory. Our proposed method predicted travel-time values more accurately than the two baseline models, and it also exhibited slower decay in performance when predicting over larger intervals. We also found that the number of station pairs has an impact on the model. When there are a sufficient number of station pairs, the model can effectively utilize the rich spatial information and achieve higher accuracy. Our approach, which incorporates spatiotemporal information, provides outputs that are more efficient and accurate compared with the traditional single-station (pair) method that only considers temporal information, suggesting that spatial information does enhance the performance of the model.

The divergence angle, as one of the important parameters of terahertz radiation, is a significant indicator of the energy distribution of the beam and its accuracy affects the signal-to-noise ratio of the system measurement. We focus on two 100 GHz continuous-wave terahertz sources with different divergence angles and propose a method for analyzing the contour envelope trend of the spot image using image processing techniques, based on the intensity distribution of terahertz radiation. The method involves constructing a conical surface based on the contour point coordinates and then iteratively obtaining the divergence angle, propagation direction, and waist position. Additionally, an uncertainty model is established based on the error source in the conical surface fitting-based divergence angle measurement experiment. Furthermore, a comparative experiment is conducted by combining the focal spot method and the knife-edge method, which verifies the feasibility and accuracy of the conical surface fitting method. The results show that the measured divergence angle of terahertz source I is 2.73 deg with an extended uncertainty of U=3. 88% (k=2), and the measured divergence angle of terahertz source II is 2.23 deg with an extended uncertainty of U=3. 90% (k=2). The measurement of terahertz sources with different divergence angles further demonstrates that the conical surface fitting method is still feasible for sources with different divergence angles. (c) 2024 Society of Photo-Optical Instrumentation Engineers (SPIE)

Aiming at the problem of the difficult segmentation of adherent images due to the not fully convex shape of peanut pods, their complex surface texture, and their diverse structures, a multimodal fusion algorithm is proposed to achieve a 2D segmentation of adherent peanut images with the assistance of 3D point clouds. Firstly, the point cloud of a running peanut is captured line by line using a line structured light imaging system, and its three-dimensional shape is obtained through splicing and combining it with a local surface-fitting algorithm to calculate a normal vector and curvature. Seed points are selected based on the principle of minimum curvature, and neighboring points are searched using the KD-Tree algorithm. The point cloud is filtered and segmented according to the normal angle and the curvature threshold until achieving the completion of the point cloud segmentation of the individual peanut, and then the two-dimensional contour of the individual peanut model is extracted by using the rolling method. The search template is established, multiscale feature matching is implemented on the adherent image to achieve the region localization, and finally, the segmentation region is optimized by an opening operation. The experimental results show that the algorithm improves the segmentation accuracy, and the segmentation accuracy reaches 96.8%.

The operational status of geophones plays a pivotal role in ensuring the accuracy and reliability of microseismic monitoring systems. However, conventional techniques used to evaluate the operational status of geophones require human intervention or significant time delays. To address this issue, we propose a method for online monitoring of geophone status using observed data obtained from a microseismic system. First, the energy features of the preprocessed observation data are extracted via wavelet packet decomposition. Subsequently, the distribution parameters of energy features are obtained through log-logistic distribution fitting. These parameters are then applied to a change-point detection model, enabling the online monitoring of seismic geophones. In addition, we select a long short-term memory network to classify the operational status of the geophones, which is trained using the obtained energy distribution data and the time-frequency characteristics of the observed data. The experimental results indicate that the model achieves an accuracy of 98.33%, surpassing the 89.58% accuracy of the support vector machine. The proposed method not only contributes to online monitoring and precise determination of the operating status of detectors, but also has enormous application potential in other fields that require monitoring and evaluating the operating status of instruments.

The method of seismic ambient noise cross-correlations (NCFs) has been demonstrated to be applicable for time offset measurement, especially in evaluating instrument clocks. Continuous recording of seismic ambient noise data makes it possible to analyze the performance of seismic instruments online. However, long-term cross-correlation and stacking calculations are required to obtain accurate travel time from ambient noise, which greatly reduces the time resolution of instrument performance detection. Therefore, we propose a travel time extraction method based on time-frequency analysis, which could obtain the travel time accurately even though the NCF has a low signal-to-noise ratio (SNR), and it is applied to the measurement of clock offsets in seismic instruments. The method combines S Transform and dictionary learning to improve the SNR of NCFs, and uses a peak extraction algorithm based on short-time Fourier transform to obtain accurate travel times. Additionally, the travel time drift of the station pair can be obtained by calculating the travel time difference between causal and acausal parts of NCFs. Using the data from ambient noise observations and the real data with time errors to verify that the proposed method can obtain accurate travel time for NCF with SNR lower than 6 and it can pick up a time offset as low as one sampling point, the detectable change is 0.046% of travel time, which is crucial for detecting weak changes in the performance of seismic instruments.

针对工件测量尺寸精度受到物距和光源强度影响的问题,开展了基于物距和光源强度反馈的测量精度补偿研究.提出一种基于系统像素当量标定值和光源强度像素数误差量串行耦合的综合补偿方法,首先根据物距变化动态调整系统像素当量标定值;然后通过不同光源强度下的标准件实测像素数,建立光源强度像素数误差量模型;最后对图像物理尺寸测量结果进行综合补偿实验.实验结果表明:综合补偿法能够有效提高尺寸测量的精度,使测量误差小于14 μm.