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.

In power transmission systems, conductor vibration may cause periodic variations in the axial load carried by composite suspension insulator strings. Such axial-load variations may induce parametric vibration, affecting the mechanical reliability of transmission lines and the safe operation of transmission networks. This study examines how excitation frequency and amplitude affect the parametric response of a reference core-rod model for an FXBZ-500/300 composite suspension insulator. The core rod is modeled as a uniform Euler–Bernoulli beam, and the resulting eigenvalue problem is solved numerically. Its lateral restraint is calculated from published conductor data and stated support assumptions. A first-mode Galerkin reduction shows that the fluctuating axial load modulates the system stiffness through a time-varying geometric-stiffness term. The resulting undamped equation takes the Mathieu form, with the primary instability region centered near $f_{\Omega }=2f_1$. At exact resonance, the linearized damped model predicts a damping-dependent load-amplitude threshold. The first six natural frequencies predicted by the analytical beam model differ by less than $0.7\%$ from those obtained from the prestressed finite-element model, indicating numerical consistency between the two idealized beam models. For the prescribed small initial disturbance, the tested cases show decay below this threshold and bounded finite-amplitude first-mode oscillations above it. As the dynamic-load amplitude increases, the calculated midpoint-displacement amplitude increases and the response peak shifts to a higher excitation frequency, consistent with stretching-induced hardening under the fixed-end-separation reference assumption. Within the adopted model, these results identify combinations of axial-load frequency and amplitude that produce pronounced core-rod vibration. Their relevance to an operating line depends on a sustained axial-load component near $2f_1$ being transmitted to the string.

Transmission tower guy wires are critical flexible tension members ensuring the stability and safe operation of overhead power transmission networks. However, these components are vulnerable to external impacts from falling rocks, ice masses, and other natural hazards, which can cause excessive deformation, anchorage loosening, and catastrophic failure. Current design standards primarily consider static loads, lacking comprehensive models for predicting dynamic impact responses. This study presents a theoretical model for predicting the peak impact response of guy wires by modeling the impact process as a point mass impacting a nonlinear spring system. Using an energy-based elastic potential method combined with cable theory, analytical solutions for axial force, displacement, and peak impact force are derived. Newton–Cotes numerical integration solves the implicit function to obtain closed-form solutions for efficient prediction. Validated through finite element simulations, deviations of peak displacement, peak impact force, and peak axial force between theoretical and numerical results are within ±4%, ±18%, and ±4%, respectively. Using the validated model, parametric studies show that increasing the inclination angle from 15° to 55° slightly reduces peak displacement by 2–4%, impact force by 1–13%, and axial force by 1–10%. Higher prestress (100–300 MPa) decreases displacement and impact force but increases axial force. Longer lengths (15–55 m) cause linear displacement growth and nonlinear force reduction. Impacts near anchorage points help control displacement risks, and impact velocity generally has a more significant influence on response characteristics than impactor mass. This model provides a scientific basis for impact-resistant design of power grid infrastructure and guidance for optimizing de-icing strategies, enhancing transmission system safety and reliability.