Wind energy is a vital pillar of modern sustainable power generation, yet wind turbine generators remain vulnerable to incipient inter-turn short-circuit (ITSC) faults in their stator windings. These faults can cause fluctuations in the output voltage, frequency, and power of wind turbines, eventually leading to overheating, equipment damage, and rising maintenance costs if not detected early. Although significant progress has been made in condition monitoring, the current methods still fall short of the robustness required for early fault diagnosis in complex operational settings. To address this gap, this study presents a novel deep learning framework that involves traditional baseline machine-learning algorithms and advanced deep network architectures to diagnose seven distinct ITSC fault types using signals from current, vibration, and axial magnetic flux sensors. Our approach is rigorously evaluated using metrics such as confusion matrices, accuracy, recall, average precision (AP), mean average precision (mAP), hypothesis testing, and feature visualization. The experimental results demonstrate that deep learning models outperform machine learning algorithms in terms of precision and stability, achieving an mAP of 99.25% in fault identification, with three-phase current signals emerging as the most reliable indicator of generator faults compared to vibration and electromagnetic data. It is recommended to combine three-phase current sensors with deep learning frameworks for the precise identification of various types of incipient ITSC faults. This study offers a robust and efficient pipeline for condition monitoring and ITSC fault diagnosis, enabling the intelligent operation of wind turbines and maintenance of their operating states. Ultimately, it contributes to providing a practical way forward in enhancing turbine reliability and lifespan.

Real-time insulator defect detection is critical for ensuring the reliability and safety of power transmission systems. However, deploying deep learning models on edge devices presents significant challenges due to limited computational resources and strict latency constraints. To address these issues, we propose YOLOLS, a lightweight and efficient detection model derived from YOLOv8n and optimized for real-time edge deployment. Specifically, YOLOLS integrates GhostConv to generate feature maps through stepwise convolution, reducing computational redundancy while preserving representational capacity. Moreover, the C2f module is restructured into a ResNet–RepConv architecture, in which convolution and Batch Normalization layers are fused during inference to reduce model complexity and enhance inference speed. To further optimize performance, a lightweight shared-convolution detection head significantly reduces parameter count and computational cost without compromising detection accuracy. Additionally, an auxiliary bounding box mechanism is incorporated into the CIoU loss function, improving both convergence speed and localization precision. Experimental validation on the CPLID dataset demonstrates that YOLOLS achieves a 42.4% reduction in parameters and a 48.1% decrease in FLOPs compared to YOLOv8n while maintaining a high mAP of 91%. Furthermore, when deployed on Jetson Orin NX, YOLOLS achieves 44.6 FPS, ensuring real-time processing capability. Compared to other lightweight YOLO variants, YOLOLS achieves a better balance between accuracy, computational efficiency, and inference speed, making it an optimal solution for real-time insulator defect detection in resource-constrained edge computing environments.

Transmission-line inspection requires a single detector to localize complete insulator strings together with much smaller broken-shell and flashover-damaged regions. These targets differ in spatial extent and in their dependence on local detail and surrounding context, which complicates cross-scale feature fusion when accurate bounding boxes are required. This study develops receptive-field-aware path aggregation (PRA) and bidirectional receptive-field-aware aggregation (BRA), combining path aggregation with receptive-field expansion and adaptive scale weighting. Compared with the path aggregation network (PANet), PRA increases mean average precision (mAP) at an intersection-over-union (IoU) threshold of 0.5 (mAP@0.5) by 2.2 percentage points and mAP averaged over IoU thresholds from 0.5 to 0.95 (mAP@0.5: 0.95) by 4.6 points; BRA produces gains of 1.5 and 4.0 points, respectively. The highest mAP@0.5:0.95 of 91.3% is jointly achieved by Swin-T-PRA + Alpha-CIoU and Swin-T-BRA + Alpha-CIoU, with corresponding mAP@0.5 values of 98.6% and 98.5% and model-only inference speeds of 48.5 and 48.3 FPS on an RTX 3080 Ti, respectively. The larger gains across stricter IoU thresholds indicate that PRA/BRA provide greater benefits when more stringent box-overlap criteria are imposed, although the present evaluation does not independently isolate the bounding-box regression mechanism.