Insulator parallel gaps in a transmission line protect the insulator from arc burn and provide protection from lightning. Detection of insulator parallel-gap faults based on aerial inspection images of transmission lines may be hindered by various factors, such as complex background, scale distortion, and occlusion. In this study, we propose a prior knowledge transfer and attention mechanism network (PKAMNet) to detect faulty insulator parallel gaps. First, the capability of PKAMNet to learn the features of different types of parallel-gap faults is improved by constructing a prior knowledge transfer model based on visual saliency. Owing to the difficulty of effectively expressing the features of the fault target due to the deviation of the shooting angle and complex background of the inspection image, we enhance the feature expression ability of the fault area by embedding the coordinate attention block. Additionally, the parameters of the detection network are simplified, and the modules are optimized to improve detection efficiency. Comparison results on 3-year aerial inspection videos of 500 kV high-voltage transmission lines show that PKAMNet has a high detection accuracy and can ameliorate the phenomenon of missed and erroneous detection caused by insufficient expression of parallel-gap faults.

Photovoltaic (PV) power generation technology is an effective approach for alleviating the current energy crisis. However, PV modules are prone to hot-spot faults and shorter service lives because of harsh long-term operating conditions. Therefore, in this study, we propose a bi-branch collaborative training algorithm based on knowledge distillation for PV hot-spot detection systems, focusing on the accuracy and computational efficiency of the algorithm. The proposed approach uses a knowledge distillation model based on a “teacher and student” collaborative training structure to improve detection accuracy while ensuring inference speed. The anchor-free YOLOX detection algorithm is used as the student network to reduce the number of network parameters. This includes a decoupled prediction head designed to improve detection performance. The teacher network first uses a HorNet-based cross-stage partial network (CSPHN)-based backbone network to enhance feature representation ability by capturing higher order spatial interactions. Second, a bi-branch multilevel feature adaptive fusion (BiMAF) module is designed to fuse multiscale features from both global and local perspectives in a parallel manner. Third, a shift contextual transformer (SCT)-based prediction head is adopted to enhance the long-range interaction between cross-scale features, thereby significantly improving the accuracy of the detection algorithm in dense scenes. Finally, comparative experiments are conducted on 13 classical detection algorithms to verify the effectiveness of the proposed detection network. The experimental results demonstrated that the proposed system can achieve fast and accurate detection of multiscale hot-spot faults under various harsh conditions, achieving an AP50 metric of 82.2%.

To address the problem of traditional pedestrian detection methods being subject to random interference from the external environment and insufficient utilization of pedestrian feature information, a novel multi-modal pedestrian detection network called MDFOaNet is proposed. The proposed detection network consists of two key components: a marine predator-based multi-scale image fusion module and a pedestrian detection module with an enhanced target visual accuracy attention model. In the fusion module, a contrast-based layered enhancement method for infrared images and a sharpness-based enhancement method for visible images are proposed for the problem of blurred pedestrian features in images. Moreover, to control the trade-off between fusion sub-layers, a dynamic image reconstruction model that relies on adaptive optimization based on marine predators is designed. Meanwhile, in the pedestrian detection module, to pay more attention to the main information in image and ignore some irrelevant information, an EVAM attention model is designed under the framework of YOLOv5s detection network, which improves the saliency of pedestrian targets and suppress the background interference. The experimental results show that compared with nine typical algorithms, the proposed algorithm can achieve accurate detection of multi-scale targets in complex environments, and is significantly superior to the compared detection algorithms in both subjective and objective evaluation indicators. The mAP and recall rates of the proposed network can reach 88.9% and 87.5%, respectively.

In view of the low levels of speed and precision associated with fault detection in transmission lines using traditional algorithms due to resource constraints, a transmission line fault target-detection method for YOLOv8 (You Only Look Once version 8) based on the Rep (Representational Pyramid) Visual Transformer and incorporating an ultra-lightweight module is proposed. First, the YOLOv8 detection network was built. In order to address the needs of feature redundancy and high levels of network computation, the Rep Visual Transformer module was introduced in the Neck part to integrate the pixel information associated with the entire image through its multi-head self-attention and enable the model to learn more global image features, thereby improving the computational speed of the model; then, a lightweight GSConv (Grouped and Separated Convolution, a combination of grouped convolution and separated convolution) convolution module was added to the Backbone and Neck to share computing resources among channels and reduce computing time and memory consumption, by which the computational cost and detection performance of the detection network were balanced, while the model remained lightweight and maintained its high precision. Secondly, the loss function Wise-IoU (Intelligent IOU) was introduced as the Bounding-Box Regression (BBR) loss function to optimize the predicted bounding boxes in these grid cells and shift them closer to the real target location, which reduced the harmful gradients caused by low-quality examples and further improved the detection precision of the algorithm. Finally, the algorithm was verified using a data set of 3500 images compiled by a power-supply inspection department over the past four years. The experimental results show that, compared with the seven classic and improved algorithms, the recall rate and average precision of the proposed algorithm were improved by 0.058 and 0.053, respectively, compared with the original YOLOv8 detection network; the floating-point operations per second decreased by 2.3; and the picture detection speed was increased to 114.9 FPS.