Search

Insulators are crucial insulation components and structural supports in power grids, playing a vital role in the transmission lines. Due to temperature fluctuations, internal stress, or damage from hail, insulators are prone to injury. Automatic detection of damaged insulators faces challenges such as diverse types, small defect targets, and complex backgrounds and shapes. Most research for detecting insulator defects has focused on a single defect type or a specific material. However, the insulators in the grid's transmission lines have different colors and materials. Various insulator defects coexist, and the existing methods have difficulty meeting the practical application requirements. Current methods suffer from low detection accuracy and mAP0.5 cannot meet application requirements. This paper proposes an improved you only look once version 7 (YOLOv7) model for multi-type insulator defect detection. First, our model replaces the spatial pyramid pooling cross stage partial network (SPPCSPC) module with the receptive filed block (RFB) module to enhance the network's feature extraction capability. Second, a coordinate attention (CA) mechanism is introduced into the head part to enhance the network's feature representation ability and to improve detection accuracy. Third, a wise intersection over union (WIoU) loss function is employed to address the low-quality samples hindering model generalization during training, thereby improving the model's overall performance. The experimental results indicate that the proposed model exhibits enhancements across various performance metrics. Specifically, there is a 1.6% advancement in mAP_0.5, a corresponding 1.6% enhancement in mAP_0.5:0.95, a 1.3% elevation in precision, and a 1% increase in recall. Moreover, the model achieves parameter reduction by 3.2 million, leading to a decrease of 2.5 GFLOPS in computational cost. Notably, there is also an improvement of 2.81 milliseconds in single-image detection speed. This improved model can detect insulator defects for diverse materials, color insulators, and partial damage shapes in complex backgrounds.

Abstract It is very challenging to accurately segment smoke images because smoke has some adverse vision characteristics, such as anomalous shapes, blurred edges, and translucency. Existing methods cannot fully focus on the texture details of anomalous shapes and blurred edges simultaneously. To solve these problems, a Dense Multi‐scale context and Asymmetric pooling Embedding Network (DMAENet) is proposed to model the smoke edge details and anomalous shapes for smoke segmentation. To capture the feature information from different scales, a Dense Multi‐scale Context Module (DMCM) is proposed to further enhance the feature representation capability of our network under the help of asymmetric convolutions. To efficiently extract features for long‐shaped objects, the authors use asymmetric pooling to propose an Asymmetric Pooling Enhancement Module (APEM). The vertical and horizontal pooling methods are responsible for enhancing features of irregular objects. Finally, a Feature Fusion Module (FFM) is designed, which accepts three inputs for improving performance. Low and high‐level features are fused by pixel‐wise summing, and then the summed feature maps are further enhanced in an attention manner. Experimental results on synthetic and real smoke datasets validate that all these modules can improve performance, and the proposed DMAENet obviously outperforms existing state‐of‐the‐art methods.

The complex terrain of China frequently leads to wildfires, which in turn pose a threat to the safe operation of power transmission lines. Studying the breakdown characteristics of air gaps under wildfire conditions is of great significance for understanding wildfire propagation mechanisms, risk assessment and management, and ecological environment protection. This paper establishes an experimental platform simulating wildfire climatic conditions and conducts experimental research on air gaps between rod–rod gaps and conductor–ground gaps. The experimental voltage types include direct current, power frequency, and standard operating waves. The impact of wildfire factors on the breakdown voltage and discharge characteristics of air gaps was obtained. The results indicate that the main factors affecting the air gap breakdown characteristics during wildfires are flame height and smoke. Flame height directly influences the gap insulation distance. Under flame bridging conditions, the maximum decrease in breakdown voltage reaches 70–80%. As the concentration of smoke increases, the degradation of insulation performance becomes more pronounced, with a reduction ranging from 20% to over 50%.

Lightning is an electrical discharge phenomenon in the atmosphere caused by charge separation in clouds, which is divided into cloud-to-ground (CG) and cloud-to-cloud (CC) lightning. In recent years, research on the characteristics of multiple-stroke ground lightning and the amplitude of lightning currents has attracted significant attention. The amplitude of lightning currents serves as fundamental data for lightning protection in power systems. Its accurate measurement is crucial for designing and safeguarding power systems. This paper obtains data from a lightning location system and analyzes the probability density distribution of lightning current amplitudes. It is found that the median of lightning currents gradually decreases with an increasing number of multiple strokes, and there is a trend in the change of lightning current steepness. As the number of strokes increases, the median value of amplitude distribution gradually decreases, while the steepness coefficient shows an increasing trend. These research findings contribute to a deeper understanding of the characteristics of lightning and provide important references for lightning prevention and disaster reduction.

Accurate cloud detection is an important step to improve the utilization rate of remote sensing (RS). However, existing cloud detection algorithms have difficulty in identifying edge clouds and broken clouds. Therefore, based on the channel data of the Himawari-8 satellite, this work proposes a method that combines the feature enhancement module with the Gaussian mixture model (GMM). First, statistical analysis using the probability density functions (PDFs) of spectral data from clouds and underlying surface pixels was conducted, selecting cluster features suitable for daytime and nighttime. Then, in this work, the Laplacian operator is introduced to enhance the spectral features of cloud edges and broken clouds. Additionally, enhanced spectral features are input into the debugged GMM model for cloud detection. Validation against visual interpretation shows promising consistency, with the proposed algorithm outperforming other methods such as RF, KNN and GMM in accuracy metrics, demonstrating its potential for high-precision cloud detection in RS images.

Fire incidents near power transmission lines pose significant safety hazards to the regular operation of the power system. Therefore, achieving fast and accurate smoke detection around power transmission lines is crucial. Due to the complexity and variability of smoke scenarios, existing smoke detection models suffer from low detection accuracy and slow detection speed. This paper proposes an improved model for smoke detection in high-voltage power transmission lines based on the improved YOLOv7-tiny. First, we construct a dataset for smoke detection in high-voltage power transmission lines. Due to the limited number of real samples, we employ a particle system to randomly generate smoke and composite it into randomly selected real scenes, effectively expanding the dataset with high quality. Next, we introduce multiple parameter-free attention modules into the YOLOv7-tiny model and replace regular convolutions in the Neck of the model with Spd-Conv (Space-to-depth Conv) to improve detection accuracy and speed. Finally, we utilize the synthesized smoke dataset as the source domain for model transfer learning. We pre-train the improved model and fine-tune it on a dataset consisting of real scenarios. Experimental results demonstrate that the proposed improved YOLOv7-tiny model achieves a 2.61% increase in mean Average Precision (mAP) for smoke detection on power transmission lines compared to the original model. The precision is improved by 2.26%, and the recall is improved by 7.25%. Compared to other object detection models, the smoke detection proposed in this paper achieves high detection accuracy and speed. Our model also improved detection accuracy on the already publicly available wildfire smoke dataset Figlib (Fire Ignition Library).

This paper investigates ionospheric response characteristics from multiple perspectives based on globally distributed GNSS data and products, ionosonde data, FORMOSAT-7/COSMIC-2 occultation data, and Swarm satellite observations caused by the total solar eclipse of 8 April 2024 across North and Central America. The results show that both GNSS-derived TEC products have detected the ionospheric TEC degradation triggered by the total solar eclipse, with the maximum degradation exceeding 10 TECU. The TEC data from nine GNSS stations in the path of the maximum eclipse reveal that the intensity of ionospheric TEC degradation is related to the spatial location, with the maximum degradation value of the ionospheric TEC being about 14~23 min behind the moment of the maximum eclipse. Additionally, a negative anomaly of foF2 with a maximum of more than 2.7 MHz is detected by ionosonde. In the eclipse region, NmF2 and hmF2 show trends of decrease and increase, with percentages of variation of 40~70% and 4~16%, respectively. The Ne profile of the Swarm-A satellite is significantly lower than the reference value during the eclipse period, with the maximum negative anomaly value reaching 11.2 × 105 el/cm3, and it failed to show the equatorial ionization anomaly.

As the speedy expansion of power gird, the traveling wave fault locator in substation may not always connect to newly added transmission lines in time, which reduces the effectiveness of fault locator. In this paper, an adjacent fault-free line based fault location scheme for newly additional line is proposed. Since the traveling wave from adjacent fault-free line results in severe difficulty for identification of traveling waves from faulty line, the time–frequency matrix is performed to screen traveling waves. Firstly, in order to remove the influence of waves from remote terminals of adjacent fault-free line, the current refraction coefficient is theoretically analyzed. Due to the negative refraction coefficient, traveling waves from remote terminals of adjacent fault-free line can be reduced by the proposed composite traveling wave. Secondly, the time–frequency matrices of composite traveling wave and original traveling wave is obtained by S-transform. Then, these matrices are applied to form ratio matrix which thoroughly eliminated waves from remote terminals of fault-free lines. Moreover, the robustness of the proposed fault location scheme is thoroughly validated by the simulation of PSCAD/EMTDC. Finally, the proposed fault location method has the great accuracy in simulation tests and field data test.

The compact design of a 500 kV quadruple-circuit transmission line can effectively reduce the line corridor area, but the height of the tower also increases, increasing the probability of suffering a lightning strike. The 500 kV quadruple-circuit transmission lines carry more energy, and because of this, lightning strikes that cause power line trips are more likely to result in large-scale power outages. Therefore, it is necessary to make an accurate assessment of the lightning performance of 500 kV quadruple-circuit transmission lines. First, simulated lightning-striking experiments were carried out on a scaled 500 kV quadruple-circuit transmission line in the laboratory, where transient voltages and currents were measured. Second, a numerical model was established with the Finite-Difference Time Domain (FDTD) method, which was then verified with the experimental results. Third, lightning surge responses of a 500 kV quadruple-circuit transmission line under near-real facility conditions are estimated with the verified FDTD model. In the simulations, influencing factors, such as the rise time of injecting current, the velocity of return-stroke current and the terrains, were taken into consideration, but not in previous lightning surge analysis with the Electromagnetic Transients Program (EMTP). Results show that insulator voltages on the same tower cross-arm are nearly identical, although the length of the cross-arm is large enough. Furthermore, it is found that the rise time and the lightning current velocity have great effects on the lightning surge response, and the terrains are less impactful but not negligible. Therefore, these factors should be considered carefully where higher accuracy lightning protection design is necessary.

After a transmission line is covered by ice in winter, ice-shedding and vibration occurs under special meteorological and external dynamic conditions, which leads to intense transmission line shaking. Transmission line ice-shedding and vibration often cause line flashover trips and outages. In January 2018, three 500 kV transmission lines, namely, the 500 kV Guanli line, the 500 kV Dushan line, and the 500 kV Guanqiao line, tripped and cut off due to ice-shedding and vibration in Anhui province, seriously threatening the safe operation of a large power grid. Current studies mainly focus on analyzing the influence factors and characteristics of line ice-shedding and investigating suppression measures, but they only analyze the correlation between each influencing factor and icing or shedding, and do not consider the coupling effects between multiple factors. In this paper, the key influencing factors and the probability distribution of transmission line ice-shedding were analyzed, and a multiple-factor coupling fault probability calculation model of line ice-shedding based on Copula function was proposed. The fault probability was calculated directly by considering multiple influence factors at the same time, which effectively overcame the error caused by multi-factor transformation in fuzzy membership degree and other methods. It provided an important decision-making basis for preventing and controlling transmission line ice-shedding faults.

Wire icing is a prevalent challenge in both industrial and scientific domains, and it is widely acknowledged that terrain and water vapor are significant contributing factors in the formation of wire icing. Consequently, the identification of terrains that are prone to inducing water vapor uplift serves as the scientific foundation for predicting ice accumulation on power lines. Yunnan Province, a mountainous province in China, features a large elevation difference. In winter, this region is prone to wire, pole and tower icing, which can affect power transmission and cause economic and property losses. Therefore, it is necessary to conduct research on the identification of typical terrain patterns in this region. In previous terrain studies, more attention has been focused on slope and aspect, watershed analysis and terrain profile analysis. When the purpose of the terrain identification is to analyze which terrains are more prone to collecting water vapor, we hope to obtain slightly larger terrain blocks and analyze the water vapor sources for different terrains in order to identify typical terrains that are conducive to icing formation. A new technology for identifying terrain patterns based on the K-means clustering method is proposed in this study to explore the typical terrain in Yunnan province. Additionally, the influences of different terrain patterns on water vapor movement are also analyzed. The results indicate that the typical terrains in Yunnan are “Valley-Air Channel”, “Topographic Uplifting”, “Ravine”, “Mountain Pass” and “Alpine Divide” patterns. The results show that the identified typical terrain is consistent with observations from satellite images, which verifies the effectiveness of this identification method. Among these five typical terrains, the “Valley-Air Channel”, the “Topographic Uplifting” and the “Mountain Pass” terrains are prone to collecting water vapor and forming ice cover. The “Alpine Divide” terrain is also prone to accumulating water vapor on both sides to form ice cover. The identified typical terrain demonstrates that typical terrain patterns near water bodies are more prone to the occurrence of wire, pole and tower icing because these areas are abundant in water vapor, and the extensive water vapor is easily condensed under the effects of terrain uplifting and cooling. In these key areas, existing wires and towers, as well as those to be constructed in the future, deserve our special attention.

Early smoke detection using Digital Image Processing technology is an important research field, which has great applications in reducing fire hazards and protecting the ecological environment. Due to the complex changes of color, shape and size of smoke with time, it is challenging to accurately recognize smoke from a given image. In addition, limited by domain shift, the trained detector is difficult to adapt to the smoke in real scenes, resulting in a sharp drop in detection performance. In order to solve this problem, an unsupervised domain adaptive smoke detection algorithm rely on Multilevel feature Cooperative Alignment and Fusion (MCAF) was proposed in this paper. Firstly, the cooperative domain alignment is performed on the features of different scales obtained by the feature extraction network to reduce the domain difference and enhance the generalization ability of the model. Secondly, multilevel feature fusion modules were embedded at different depths of the network to enhance the representation ability of small targets. The proposed method is evaluated on multiple datasets, and the results show the effectiveness of the method.

A new meta-heuristic algorithm called like-attracts-like optimizer (LALO) is proposed in this article. It is inspired by the fact that an excellent person (i.e., a high-quality solution) easily attracts like-minded people to approach him or her. This LALO algorithm is an important inspiration for video robotics cluster control. First, the searching individuals are dynamically divided into multiple clusters by a growing neural gas network according to their positions, in which the topological relations between different clusters can also be determined. Second, each individual will approach a better individual from its superordinate cluster and the adjacent clusters. The performance of LALO is evaluated based on unimodal benchmark functions compared with various well-known meta-heuristic algorithms, which reveals that it is competitive for some optimizations.

The safe operation of high-voltage transmission lines ensures the power grid’s security. Various foreign objects attached to the transmission lines, such as balloons, kites and nesting birds, can significantly affect the safe and stable operation of high-voltage transmission lines. With the advancement of computer vision technology, periodic automatic inspection of foreign objects is efficient and necessary. Existing detection methods have low accuracy because foreign objects attached to the transmission lines are complex, including occlusions, diverse object types, significant scale variations, and complex backgrounds. In response to the practical needs of the Yunnan Branch of China Southern Power Grid Co., Ltd., this paper proposes an improved YOLOv8m-based model for detecting foreign objects on transmission lines. Experiments are conducted on a dataset collected from Yunnan Power Grid. The proposed model enhances the original YOLOv8m by incorporating a Global Attention Module (GAM) into the backbone to focus on occluded foreign objects, replacing the SPPF module with the SPPCSPC module to augment the model’s multiscale feature extraction capability, and introducing the Focal-EIoU loss function to address the issue of high- and low-quality sample imbalances. These improvements accelerate model convergence and enhance detection accuracy. The experimental results demonstrate that our proposed model achieves a 2.7% increase in mAP_0.5, a 4% increase in mAP_0.5:0.95, and a 6% increase in recall.

Electronic current transformer (ECT) is responsible for collecting and transmitting measurement, protection, and control signals of substations, which is regarded as a critical device for the connection of primary and secondary equipment in an intelligent substation. Naturally, a well-designed ECT is essential to ensure a safe and stable operation of power systems. This paper attempts to provide a detailed discussion and analysis of the principle, structure and characteristic of ECT, which mainly focuses on the following three aspects, i.e., steady-state accuracy, frequency response test and transient spread characteristics test. For a comprehensive performance test, a laboratory test platform is established in intelligent substations to simulate actual operation condition of ECT and further implement error accuracy analysis, frequency response test, transient response test and long-term test. Simulation results show the influence of ECT on measurement and protection accuracy, which can provide theoretical basis and evaluation criteria for further improvement of digital measurement performance and large-scale application in intelligent substations.

With the steady progress of the intelligent development of power systems, as well as the higher demand for power supply reliability. It is essential to achieve the effective monitoring of substations 24 h a day. The vigorous development of deep learning network brings strong theoretical and technical support to the unmanned and intelligent construction of the substation. To identify the on/off state of the isolation switch in the substation robot inspection image, this paper proposes a method for identifying the isolation switch state of YOLOv4 (You Only Look Once V4) network based on transfer learning. Firstly, for the insufficient number of samples, transfer learning is introduced, and the network feature extraction layer is pre-trained by using public data sets. Secondly, images of isolation switch are obtained by a fixed camera and inspection robot in the substation, and data set of isolation switch is constructed. Finally, the isolation switch data set is used to train the YOLOv4 network. The test results show that compared with YOLOv3 and YOLOv4, the network can improve the identification precision of the isolation switch.

As the core part of the power grid, the substation must carry out regular patrol inspection to ensure the normal operation of the equipment. In recent years, with the continuous increase of load, the area of the substation has gradually expanded, which makes its operating environment more complex, the intensity of patrol inspection has increased, and the manual patrol inspection method cannot meet the needs of patrol inspection. Therefore, it is necessary to introduce high-tech equipment and use patrol inspection robots to conduct regular patrol inspection, which inevitably leads to a large number of different working environments with matching standards and regulations for different equipment. In order to facilitate the staff to implement the delivery of robots matching the standards for the projects in the corresponding actual environment, the concept of knowledge atlas has been introduced and revolutionarily applied to the multi-source fusion of substation patrol robots, which not only more clearly analyzes the relationship between multi-source data, which can also make the operation and application in practical projects more convenient. This paper creatively uses the method of knowledge graph to summarize and analyze the multi-source fusion data of the field patrol robot in different working environments of a substation in Southwest China, and provides the research direction of this field and a one-stop manual for potential researchers and engineers in the future.

In recent years, the frequent fouling accidents have posed a serious threat to people’s life and property safety. Owing to the wide distribution of pollution sources and variable meteorological factors, it is a very time-consuming and labor-intensive task to map the pollution distribution using traditional methods. In this work, a study on the mapping of pollution distribution based on satellite remote sensing is carried out in Yunnan Province, China, as an example. Several machine learning methods (e.g. K-nearest neighbor, support vector machine) are used to analyze the effects of conditions such as multiple air pollution and meteorological data on pollution distribution map levels. The results indicate that the ensemble learning model has the highest accuracy of 72.32% in this application. The new pollution distribution map using this classifier has 5,506 more pixels in the most severe pollution level than the traditional map. Last, the remote sensing-based map and the manual measurement-based map were combined with corresponding experience weight to obtain a weighted pollution distribution map.

Automated inspection using unmanned aerial vehicles (UAVs) is an essential means to ensure safe operations of the power grid. Defect detection for antivibration hammers on transmission lines in inspection imagery is one of the critical tasks for automated UAV inspection. It needs a machine interpretation system to automatically detect numerous inspection images. In this paper, a high-efficiency model based on Cascade RCNN (region-convolutional neural network) is proposed to detect antivibration hammer defects with reduced costs and speedier response, which applies in energy-efficient transmission line inspection systems. Firstly, to reduce computational costs, this study modifies the Cascade RCNN with a probabilistic interpretation to achieve the best trade-off between the inference time and average precision. Secondly, an antivibration hammer defect detector (AVHDD) model is proposed that uses a deep layer aggregation-based feature extraction network and a highly effective weighted bidirectional feature fusion network to replace the original ResNet and FPN on the modified Cascade RCNN to further enhance the model performance. Finally, a fine classification (FC) scheme for the types of antivibration hammer defects is proposed based on defect features to rationalize the model. The AVHDD reached an experimental mAP of 97.24% when IoU = 0.75, which is 2.93% higher than the original Cascade RCNN, and the defect recall was 98.9% while also significantly improving the inference speed. Moreover, the experimental results indicate that the overall performance of the proposed model is superior to typical models, confirming its suitability for energy-efficient transmission line inspection systems.

Non-rigid point set and image registration are key problems in plenty of computer vision and pattern recognition tasks. Typically, the non-rigid registration can be formulated as an optimization problem. However, registration accuracy is limited by local optimum. To solve this problem, we propose a method with global to local transformation for non-rigid point sets registration and it also can be used to infrared (IR) and visible (VIS) image registration. Firstly, an objective function based on Gaussian fields is designed to make a problem of non-rigid registration transform into an optimization problem. A global transformation model, which can describe the regular pattern of non-linear deformation between point sets, is then proposed to achieve coarse registration in global scale. Finally, with the results of coarse registration as initial value, a local transformation model is employed to implement fine registration by using local feature. Meanwhile, the optimal global and local transformation models estimated from edge points of IR and VIS image pairs are used to achieve non-rigid image registration. The qualitative and quantitative comparisons demonstrate that the proposed method has good performance under various types of distortions. Moreover, our method can also produce accurate results of IR and VIS image registration.

With the widespread adoption of service-oriented architectures (SOA), services with the same functionality but the different Quality of Service (QoS) are proliferating, which is challenging the ability of users to build high-quality services. It is often costly for users to evaluate the QoS of all feasible services; therefore, it is necessary to investigate QoS prediction algorithms to help users find services that meet their needs. In this paper, we propose a QoS prediction algorithm called the MFDK model, which is able to fill in historical sparse QoS values by a non-negative matrix decomposition algorithm and predict future QoS values by a deep neural network. In addition, this model uses a Kalman filter algorithm to correct the model prediction values with real-time QoS observations to reduce its prediction error. Through extensive simulation experiments on the WS-DREAM dataset, we analytically validate that the MFDK model has better prediction accuracy compared to the baseline model, and it can maintain good prediction results under different tensor densities and observation densities. We further demonstrate the rationality of our proposed model and its prediction performance through model ablation experiments and parameter tuning experiments.

The flashover performance of insulators is the foundation of the calculation of precise lightning protection. The material and structure of an insulator affect its flashover performance but the research on the differences in the flashover performance of different insulators and the analysis of the reason for these differences have not been clear. In this paper, lightning impulse flashover tests of two types of insulators (composite and glass insulators) were carried out in the National Engineering Laboratory (Kunming) for Ultrahigh-voltage Engineering Technology. In these tests, the insulators were hung in a 110 kV double-circuit transmission tower and a standard lightning impulse (1.2/50 μs) was applied to the insulators. The discharge path and the 50% impulse flashover voltage of the insulators were recorded. At the same time, the electric fields in the vicinity of different insulators were calculated using the finite element method with COMSOL Multiphysics. The electric field distribution and the uneven coefficient were analyzed. Combined with the flashover test data and the electric field simulation, the relationship between the flashover performance (discharge path and 50% impulse flashover voltage) and the electric field (electric field distribution and uneven coefficient) is found. The simulation results are in accordance with the test data and the conclusions can provide useful references for the external insulation design of transmission lines and the optimization of insulators.