Explore the vast range of titles available.

High-voltage switchgear plays a crucial role in power systems, and its reliability is vital for the safety and stability of the grid. One major issue affecting high-voltage switchgear operation is the occurrence of discharge phenomena, which can damage the equipment and reduce grid efficiency. A common cause of discharge is metallic foreign objects inside the equipment. Detecting and removing these objects is key to preventing such discharges. This paper proposes a detection algorithm based on YOLOv5, enhanced by the integration of a Residual Efficient Convolutional Block Attention Module (RECBAM). The introduction of RECBAM into the YOLOv5 architecture improves the network’s feature extraction performance and the accuracy of detecting metallic foreign objects. Experimental results demonstrate that the proposed method achieves an average detection accuracy of 95.20%, which is 1.5% higher than the original YOLOv5, and the recall rate improves by 1.2%. Visualization results further show the superior performance of the proposed approach in metallic foreign object detection tasks.

The safety of belt conveyor operation is of great importance during coal conveyance. This paper proposes a multi-task-based GSSA-YOLOM algorithm for monitoring the state of belt conveyors, which utilizes segmentation head to detect foreign objects and belt deviation, thereby balancing the trade-offs among multiple tasks. The detection neck is responsible for multi-scale feature fusion by incorporating the Asymptotic Feature Pyramid Network (AFPN) to achieve enhanced spatial perception. Then, Groupwise Separable Convolution (GSConv) is further introduced to simplify the network architecture, reducing computational complexity while maintaining sufficient detection accuracy for edge device deployment. Moreover, the SlideLoss and Soft-NMS functions are integrated to reduce the rate of false positives and missed detections. Comparison experiments were conducted, and the results indicate that the proposed GSSA-YOLOM model can improve mAP@50 by 3.4% compared with the baseline model while reducing the number of parameters by 27%, thereby satisfying coal mine safety monitoring requirements.

Object identification has always been a difficult issue in the field of computer vision, where the objective is to identify the location and kind of particular objects in images. Convolutional neural networks have achieved significant advancements in accuracy and speed, and this is mostly due to their quick evolution. However, in practical applications, the detection of small objects is still an unresolved problem. Due to the sparse pixel information of small target, it is relatively difficult to extract small target features effectively. Although many methods have made significant progress, the detection rate and recall rate both have space for improvement under various conditions. This study suggests a small object detection technique that utilizes feature map weight self allocation and single-stage detection model FCOS to further increase the recall and accuracy of small object detection., which is used in aerial photography dataset and airport foreign object dataset respectively. The average detection accuracy and recall of the technique are roughly 10% higher than the baseline, as shown by the quantitative and qualitative trial findings, demonstrating the effectiveness of the adaptive feature fusion approach.

Wireless power transfer (WPT) systems for electric vehicles require reliable foreign object detection (FOD) mechanisms both during and prior to power transfer to ensure operational safety and efficiency. The primary purpose of this study was to develop a foreign object detection system to ensure that no objects are present in the area of magnetic coupling (between primary and secondary coils) prior to initiating power transfer. Conventional FOD techniques based on impedance, visual light, or thermal monitoring provide limited spatial information and are sensitive to coil misalignment. This paper proposes a machine learning-based FOD approach using a planar Magnetic Inductance Tomography (MIT) sensor array that enables spatial electromagnetic sensing for early detection and localisation of conductive foreign objects. A dataset comprising 17,800 measurement frames was collected using a custom STM32-based data acquisition system in the absence of (prior to) power transfer. Likewise, a dataset comprising 300 sets of measurement frames was collected during power transfer, in which each frame contains 120 electromagnetic sensor readings. This capture methodology coincides with the detection requirements of live WPT systems. Four classification models, including Random Forest, Support Vector Machine, XGBoost, and Multi-Layer Perceptron, were evaluated. To enhance robustness against sensor drift and environmental variations, feature-engineering techniques incorporating statistical, temporal, frequency-domain, and derivative-based features were developed. Experimental results demonstrate high detection accuracy under both controlled and real-world conditions. The proposed approach demonstrates the feasibility of integrating machine learning-based MIT sensing into wireless EV charging infrastructure for reliable foreign object detection.

Foreign object detection in manufacturing processes based on machine vision remains a challenge. The vastly different foreign objects and the complex background, as well as the scarcity of images with foreign objects constrain the application of traditional and deep learning methods, respectively. This paper discusses a novel method for intelligent foreign object detection and automatic data generation. A cascaded convolutional neural network to detect foreign objects on the surface of the tobacco pack is proposed. The cascaded network transforms the inspection into a two-stage YOLO based object detection, consisting of the tobacco pack localization and the foreign object detection. To address the scarcity of images with foreign objects, several data augmentation methods are introduced to avoid overfitting. Furthermore, a data generation methodology based on homography transformation and image fusion is developed to generate synthetic images with foreign objects. Models trained using synthetic images generated by this method show superior performance, which presents a viable approach to detecting newly introduced foreign objects. Extensive experimental results and comparisons on the tobacco pack foreign object dataset with several state-of-the-art methods demonstrate the effectiveness and superiority of the proposed method. The proposed cascaded foreign object detection network and synthetic data generation methodology have the potential for widespread applications.

Suspended foreign objects on transmission lines will shorten the discharge distance, easily leading to phase-to-ground or phase-to-phase short circuits, which induces outage accidents. Foreign objects are small and difficult to identify, resulting in low detection accuracy. An improved foreign object detection method based on Swin Transformer V2 and YOLOX (ST2Rep–YOLOX) is proposed. First, the feature extraction layer ST2CSP constructed by Swin Transformer V2 is used in the original backbone network to extract global and local features. Secondly, hybrid spatial pyramid pooling (HSPP) is designed to enlarge the receptive field and retain more feature information. Then, Re-param VGG block (RepVGGBlock) is introduced to replace all 3 × 3 convolutions in the network to deepen the network and improve feature extraction capabilities. Finally, experiments are carried out on the transmission lines foreign object image dataset, which was obtained using unmanned aerial vehicles (UAVs). The experimental results show that the average accuracy of the ST2Rep–YOLOX method can reach 96.7%, which is 4.4% higher than that of YOLOX. The accuracy of the nest, kite, and balloon increased by 9.3%, 15.4%, and 9.6%, and the recall increased by 6.5%, 9.4%, and 2.5%, respectively. This method has high detection accuracy, which provides an important reference for foreign object detection in transmission lines.

In order to eliminate the potential safety hazard that arises when metal foreign objects intervene in the wireless charging area of electric vehicles, this paper proposes that a metal foreign object detection method be applied to the wireless charging system of electric vehicles based on the optimal design of the array detection coil. Firstly, the equivalent circuit model of the metal foreign object detection system is established, then the principle of the foreign object detection system is analyzed, and the scale factor β is introduced as the optimization index of the detection coil. Secondly, the change of the scale factor β with the circuit parameters is analyzed and the appropriate circuit parameters are compared and selected. Thirdly, on the basis of the planar square spiral coil, Ansys Maxwell finite element simulation software is used to optimize its structural parameters, combination mode, and resonant circuit, as well as design the anti-series and anti-parallel enhanced detection coil sets with the decoupling and elimination of detection blind spots. Finally, the feasibility of the proposed detection method of metal foreign objects is verified by experiments. The results show that the two array detection coil sets can detect small-sized common metal foreign objects such as paper clips and the proposed double-layer reinforced structure can significantly improve the detection sensitivity of the system.

This research aims to overcome three major challenges in foreign object detection on power transmission lines: data scarcity, background noise, and high computational costs. In the improved YOLOv8 algorithm, the newly introduced lightweight GSCDown (Ghost Shuffle Channel Downsampling) module effectively captures subtle image features by combining 1 × 1 convolution and GSConv technology, thereby enhancing detection accuracy. CSPBlock (Cross-Stage Partial Block) fusion enhances the model’s accuracy and stability by strengthening feature expression and spatial perception while maintaining the algorithm’s lightweight nature and effectively mitigating the issue of vanishing gradients, making it suitable for efficient foreign object detection in complex power line environments. Additionally, PAM (pooling attention mechanism) effectively distinguishes between background and target without adding extra parameters, maintaining high accuracy even in the presence of background noise. Furthermore, AIGC (AI-generated content) technology is leveraged to produce high-quality images for training data augmentation, and lossless feature distillation ensures higher detection accuracy and reduces false positives. In conclusion, the improved architecture reduces the parameter count by 18% while improving the mAP@0.5 metric by a margin of 5.5 points when compared to YOLOv8n. Compared to state-of-the-art real-time object detection frameworks, our research demonstrates significant advantages in both model accuracy and parameter size.

Research on coal foreign object detection based on deep learning is of great significance to safe, efficient, and green production of coal mines. However, the foreign object image dataset is scarce due to collection conditions, which brings an enormous challenge to coal foreign object detection. To achieve augmentation of foreign object datasets, a high-quality coal foreign object image generation method based on improved StyleGAN is proposed. Firstly, the dual self-attention module is introduced into the generator to strengthen the long-distance dependence of features between spatial and channel, refine the details of the generated images, accurately distinguish the front background information, and improve the quality of the generated images. Secondly, the depthwise separable convolution is introduced into the discriminator to solve the problem of low efficiency caused by the large number of parameters of multi-stage convolutional networks, to realize the lightweight model, and to accelerate the training speed. Experimental results show that the improved model has significant advantages over several classical GANS and original StyleGAN in terms of quality and diversity of the generated images, with an average improvement of 2.52 in IS and a decrease of 5.80 in FID for each category. As for the model complexity, the parameters and training time of the improved model are reduced to 44.6% and 58.8% of the original model without affecting the generated images quality. Finally, the results of applying different data augmentation methods to the foreign object detection task show that our image generation method is more effective than the traditional methods, and that, under the optimal conditions, it improves APbox by 5.8% and APmask by 4.5%.

Wireless power transfer technology is being widely used in electric vehicle wireless-charging applications, and foreign object detection (FOD) is an important module that is needed to satisfy the transmission and safety requirements. FOD mostly includes two key parts: metal object detection (MOD) and living object detection (LOD), which should be implemented during the charging process. In this paper, equivalent circuit models of a metal object and a living object are proposed, and the FOD methods are reviewed and analyzed within a unified framework based on the proposed FOD models. A comparison of these detection methods and future challenges is also discussed. Based on these analyses, detection methods that employ an additional circuit for detection are recommended for FOD in electric vehicle wireless-charging applications.