Safety monitoring of power operations in substations is crucial for accident prevention. However, traditional methods such as wearable devices and video surveillance suffer from limitations including high costs and reliance on lighting conditions. A solution that integrates millimeter-wave radar with deep learning ensures operational compliance by high precision gesture detection. This paper proposes RoPEHAR, a human activity recognition system based on millimeter-wave radar, which combines traditional transformer architecture with rotary positional encoding, specifically designed for human posture recognition in indoor industrial environments. To reduce interference from the coupling of human and instrument signals and noise in electrical scenarios, RoPEHAR introduces a hybrid filtering pipeline that combines hierarchical SNR denoising with enhanced DBSCAN clustering to accurately segment the point cloud data of limbs and instruments. The core innovation lies in the introduction of a spatiotemporal Informer, Roformer. It enhances the 3D-space vector information about data points through dynamic rotary positional encoding. This system effectively models limb motion trajectories. Experiments demonstrate that RoPEHAR achieves high-precision performance, reaching an accuracy of 95.8%, enabling real-time and reliable activity recognition for substation safety monitoring.

Accurate and scalable indoor localization for unmanned aerial vehicles (UAVs) is essential for Internet of Things (IoT) applications such as autonomous logistics, infrastructure inspection, and emergency response in GPS-denied environments. However, traditional methods often struggle with cost, deployment complexity, and sensitivity to environmental dynamics, limiting their practicality for large-scale IoT scenarios. This paper presents a method in which Channel State Information (CSI) from low-cost IoT sensors enables robust, device-free 3D UAV localization while optimizing accuracy, sensor adaptability, and data efficiency. We propose CiUAV, leveraging CSI captured by ESP32-S3 sensors, with a Robust CSI Signal Enhancement (RCSE) framework integrating Dynamic AGC Compensation (DAC) and Adaptive Noise Suppression and Outlier Removal (ANSOR), alongside a Sensor-in-Sample (SiS) multi-objective optimization model for adaptive multi-sensor fusion. Experimental evaluations in realistic indoor settings achieve a 3D root mean squared error (RMSE) of 0.2659 meters, outperforming baselines by up to 35% in accuracy and 50% in data efficiency. CiUAV offers a lightweight, scalable, and infrastructure-compatible solution for future IoT-enabled UAV systems.

Ensuring personnel safety during power station inspections is a critical yet challenging task due to inherent hazards in such environments. Traditional monitoring methods, including wearable devices and video surveillance, suffer from user discomfort, limited visibility, and high deployment costs. To overcome these limitations, this article proposes PowerHAR, a device-free framework for recognizing power inspection activities based on WiFi channel state information (CSI) acquired from custom-designed ESP32 internet of things (IoT) sensors. PowerHAR introduces a spatio-temporal feature distribution-based power operation recognition network, comprising a transformer-based preprocessing module capable of effectively handling variable-length CSI sequences, and a spatio-temporal extraction module that integrates convolutional operations with multihead self-attention mechanisms for comprehensive feature fusion. By leveraging mutual CSI sensing among distributed sensors, PowerHAR provides robust and accurate recognition of power inspection activities without requiring additional hardware infrastructure. Experimental validation demonstrates that PowerHAR significantly surpasses existing baseline methods, confirming its high reliability and practicality in safety-critical industrial scenarios.

Balanced in-core power levels in nuclear power plants (NPPs) are critical for safety, whereas power tilt disrupts this balance, reducing safety margins and posing risks. Early warning for power tilt offers an effective way of optimizing monitoring. Due to abnormal-sample scarcity and security concerns, common data-driven models train on the simulated data generated by simulators. However, achieving a satisfactory effect in practices is difficult because simulators imperfectly emulate reality. Thus, we propose a power tilt-oriented early warning method called simulation–reality spatial–temporal model (SR-STM). Motivated by the physical model in NPPs, a knowledge-guided hierarchical graph is designed to characterize spatial correlations among local power levels for SR-STM’s input. The SR-STM uses a lightweight spatial–temporal network (LST-Net) as a feature extractor, balancing precision, and efficiency. To bridge sim-real interdomain discrepancies, SR-STM utilizes node-alignment adversarial learning (NAAL) for fine weight tuning in subdomain, and eigenvalue-based scale alignment (ESA) for sim-real feature proximity. Forecasting local power levels using the SR-STM, dynamic metrics and alarm limits are calculated and compared to perform the early warning task. The online experimental prototype verifies that SR-STM surpasses various state-of-the-art methods in terms of early warning and sim-real cross-domain tasks.

Neutron detectors in nuclear power plants (NPPs) are critical for system stability, yet their malfunctions may lead to false alerts and misdiagnoses. Multidetectors deployed in diverse positions vary with the nuclear reactor states contained spatial-temporal variations of neutron fluxes. Existing methods seldom concurrently consider intricate spatial-temporal correlations and gradual state variations among detectors. This study proposes a detector-oriented fault detection and isolation method named the spatial-temporal state adaptation model (ST-SAM). The method introduces a local-global spatial-temporal network that captures the potential interdependencies within the detector topology. To minimize cross-state discrepancies in reactors, ST-SAM integrates three submodules: a signal reconstructor to enhance the specific-state variation representation; a correlation alignment to mitigate interstate feature discrepancies; and an adversarial discriminator to extract spatial-temporal state-invariant features. Leveraging the parallel detection strategy, ST-SAM effectively detects and isolates faulty detectors, preventing fault propagation on subsequent diagnosis. Experiments on ex-core and in-core neutron detectors in real-world NPPs with simulated faults verify that the ST-SAM outperforms various state-of-the-art methods in terms of signal reconstruction and fault detection.

Safety monitoring of power operations in power stations is crucial for preventing accidents and ensuring stable power supply. However, conventional methods such as wearable devices and video surveillance have limitations such as high cost, dependence on light, and visual blind spots. WiFi-based human pose estimation is a suitable method for monitoring power operations due to its low cost, device-free, and robustness to various illumination conditions. In this paper, a novel Channel State Information (CSI)-based pose estimation framework, namely PowerSkel, is developed to address these challenges. PowerSkel utilizes self-developed CSI sensors to form a mutual sensing network and constructs a CSI acquisition scheme specialized for power scenarios. It significantly reduces the deployment cost and complexity compared to the existing solutions. To reduce interference with CSI in the electricity scenario, a sparse adaptive filtering algorithm is designed to preprocess the CSI. CKDformer, a knowledge distillation network based on collaborative learning and self-attention, is proposed to extract the features from CSI and establish the mapping relationship between CSI and keypoints. The experiments are conducted in a real-world power station, and the results show that the PowerSkel achieves high performance with a PCK@50 of 96.27%, and realizes a significant visualization on pose estimation, even in dark environments. Our work provides a novel low-cost and high-precision pose estimation solution for power operation.

For intelligent inspection of power lines, a core task is to detect components in aerial images. Currently, deep supervised learning, a data-hungry paradigm, has attracted great attention. However, considering real-world scenarios, labeled data are usually limited, and the utilization of abundant unlabeled data is rarely investigated in this field. This study deploys a pretrained model for power line component detection based on a self-supervised pretraining approach, which exploits useful information from unannotated data. Concretely, we design a new masking strategy based on the structural characteristic of power lines to guide the pretraining process with meaningful semantic content. Meanwhile, a Siamese architecture is proposed to extract complete global features by using dual reconstruction with semantic targets provided by the proposed masking strategy. Then, the knowledge distillation is utilized to enable the pretrained model to learn both domain-specific and general representations. Moreover, a feature pyramid mechanism is adopted to capture multiscale features, which can benefit the detection task. Experimental results show that the proposed approach can successfully improve the performance of a variety of detection frameworks for power line components, and outperforms other self-supervised pretraining methods.

The concentration of dissolved gases in transformer oil can be utilized to diagnose faults in transformers. However, substandard online monitoring data may lead to inaccurate fault diagnosis outcomes, resulting in severe repercussions. Hence, this study presents a novel approach for anomaly detection in dissolved gas online monitoring data in transformer oil using stacking ensemble learning. Firstly, a sliding time window is employed to preprocess the monitoring data and generate a dataset consisting of time series monitoring data. Subsequently, evaluation metrics and diversity measures are applied to select distinct base learners and a meta-learner for the stacking model. This approach amalgamates the strengths and disparities of various learners. Lastly, comparative analysis of case studies demonstrates the effectiveness of the proposed method in distinguishing different types of anomalies in dissolved gas online monitoring data, exhibiting superior performance in terms of accuracy, F1 score, and area under curve(AUC).

Low-resolution infrared-based human activity recognition (HAR) attracted enormous interests due to its low cost and private. In this article, a novel semi-supervised cross-domain neural network (SCDNN) based on$8\times8$low-resolution infrared sensor is proposed for accurately identifying human activity despite changes in the environment at a low cost. The SCDNN consists of feature extractor, domain discriminator, and label classifier. In the feature extractor, the unlabeled and minimal labeled target domain data are trained for domain adaptation to achieve a mapping of the source domain and target domain data. The domain discriminator employs the unsupervised learning to migrate data from the source domain to the target domain. The label classifier obtained from training the source domain data improves the recognition of target domain activities due to the semi-supervised learning utilized in training the target domain data. Experimental results show that the proposed method achieves 92.12% accuracy for recognition of activities in the target domain by migrating the source and target domains. The proposed approach adapts superior to cross-domain scenarios compared to the existing deep learning methods, and it provides a low cost yet highly adaptable solution for cross-domain scenarios.

Detecting the components in aerial images is a crucial task in automatic visual inspection for power lines. Currently, deep learning models guided by external knowledge have achieved promising performances compared to directly applying the benchmark detectors. However, the component relationship, as human commonsense knowledge for object reasoning, is rarely investigated in this field. This study presents a graph-based relation guided network for power line component detection, which exploits correlations of regions, images, and categories. The visual relation module is employed to learn region-to-region relationship and enhance the visual features of each proposal that may contain components. Meanwhile, two guidance modules are proposed to capture image-to-region correlation and distinctively facilitate the category classification and position regression, which has not been considered in previous methods. Moreover, the category graphs built in these two modules are able to explore category-to-category dependencies that can further promote the network ability. Experimental results demonstrate that the proposed method can achieve more accurate and reasonable component detection compared to previous methods, which verifies the effectiveness of the proposed model incorporated with relation knowledge.

Abstract Digital imaging and image‐processing techniques have revolutionized the way of power line inspection in recent years. Massive images are captured and utilized for further processing to maintain the reliability, safety, and sustainability of power transmission. For power line inspection, the component region in the delivered images is required to be centered, large, and clear enough. In this paper, a component‐oriented image quality assessment method is proposed to automatically predict image quality according to the demand of power line inspection. The proposed method considers two factors: spatial characteristic evaluation and sharpness evaluation. The spatial characteristic evaluation utilizes YOLOv3 to evaluate whether the component region is sufficiently centered and large, which enables the observer to quickly find the target. The sharpness evaluation employs ResNet to evaluate the clarity of component and makes the condition monitoring more accurately. For final quality assessment, a multi‐stage filtering strategy is presented to aggregate these two factors and obtain high quality inspection images. The experimental results indicate that the high‐quality images can be accurately identified to satisfy the requirements of power line inspection. The proposed quality assessment method enhances the efficiency for further data analysis of aerial images.

Aiming at measuring tree barriers in transmission line corridors, a binocular vision ranging method is proposed to measure the distance between the transmission lines and trees. Based on the principle of binocular vision ranging, the binocular camera is calibrated using a marked checkerboard as a calibration board. Then, the SAD region matching algorithm is applied to the preprocessed images, and the disparity map is acquired. Finally, the distance between the tree and the transmission line can obtain according to the three-dimensional coordinate information of two target points. The results of the experiments show that the proposed binocular vision-based method can achieve the measurement error within ±30cm. The precision can meet the need of measuring the distance between trees and the transmission lines, which is an effective way for warning tree barriers in transmission line corridors.

Intelligent inspection in the substation transformer using optical character recognizer has been developing rapidly. Character segmentation from the text line of data plate is an important step for localization and recognition of electrical equipment. However, on-site character segmentation is challenging if the data plate contains multiple languages, especially when the width between Chinese and non-Chinese character differs significantly and the complex environments cause the light reflection and fading. This paper proposes a new method, based on analyzing the connected component and Chinese character's structure, to segment characters from multi-language data plate of substations. The proposed method uses the combination of the HSV color space and multi-scale MSRCP to reduce the effect of illumination and complex background. The proposed method utilized the width of each kind character, the interval between characters and the relationship within the left-right structure Chinese character to improve the segmentation accuracy. Experimental results show that the text lines from the data plate in substation transformer, including Chinese, English, Roman numerals, Arabic numerals and symbols, can be segmented correctly. Results show that the proposed method outperforms two existing character segmentation methods and achieves 99.4% precision in the multi-language data plate dataset.