Kun Zhang

Unmanned Aerial Vehicle (UAV) inspection of overhead transmission line composite insulators allows for the rapid acquisition of large amounts of infrared images of insulators in service, which can be used to analyse their operational status. However, efficiently and accurately analysing these images remains a challenge. This paper proposes a temperature extraction scheme based on Oriented Bounding Boxes (OBB) detection results for the intelligent extraction of the core rod sheath temperature of composite insulators on overhead lines. First, a You Only Look Once version 11 (YOLOv11) model is established to detect OBB for composite insulators in the infrared images, achieving a mAP50:95 of 91.5%. Based on the detection results, an insulation central axis localisation scheme that integrates candidate lines and skeleton extraction is proposed, which accurately locates the central axis of the composite insulator in the image. Finally, using filtering and extremum extraction methods, the temperature points along the central axis of the core rod are accurately extracted. The proposed temperature extraction scheme reduces the probability of abnormal temperature occurrences in the test set to 4.9%, nearly matching the effect of manual temperature extraction. This provides an effective framework for the intelligent processing of infrared inspection images of overhead line composite insulators.

With the widespread deployment of high-voltage and ultra-high-voltage transmission lines, composite insulators play a vital role in modern power systems. However, prolonged service leads to material aging, and the current lack of standardized, quantitative methods for evaluating silicone rubber degradation poses significant challenges for condition-based maintenance. To address this measurement gap, we propose a novel aging assessment framework that integrates Fourier Transform Infrared (FTIR) spectroscopy with a measurement-oriented ensemble learning model. FTIR is utilized to extract absorbance peak areas from multiple aging-sensitive functional groups, forming the basis for quantitative evaluation. This work establishes a measurement-driven framework for aging assessment, supported by information-theoretic feature selection to enhance spectral relevance. The dataset is augmented to 4847 samples using linear interpolation to improve generalization. The proposed model employs k-nearest neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF), and Gradient-Boosting Decision Tree (GBDT) within a two-tier ensemble architecture featuring dynamic weight allocation and a class-balanced weighted cross-entropy loss. The model achieves 96.17% accuracy and demonstrates strong robustness under noise and anomaly disturbances. SHAP analysis confirms the resistance to overfitting. This work provides a scalable and reliable method for assessing silicone rubber aging, contributing to the development of intelligent, data-driven diagnostic tools for electrical insulation systems.

The surface ageing of silicone rubber composite insulators, widely used in power systems, poses significant challenges. This study integrates Fourier transform infrared (FTIR) spectroscopy with machine learning to evaluate ageing states and explore underlying mechanisms under various environmental conditions. A dataset covering light, medium, and severe ageing was built through FTIR experiments, spectral feature extraction, and data augmentation. An ensemble learning model achieved a classification accuracy of 95.42%. SHapley Additive exPlanations (SHAP) analysis indicated that silicon–oxygen backbones, silylmethyl groups, and hydroxyl groups are key to the ageing process. The silicon–oxygen backbone is dominant in initial oxidation and cross‐linking, whereas silylmethyl group reactions occur later. Hydroxyl group changes are complex and strongly environment‐dependent during severe ageing. The model was also applied to naturally aged samples from Tibet and Inner Mongolia, showing strong classification performance and revealing clear regional differences. These findings are valuable for assessing surface ageing, analysing ageing mechanisms and developing grading standards for composite insulators.

Composite insulators are critical components in power systems, and the accurate assessment of their aging state is essential for ensuring the safe operation of the power grid. In this study, a large amount of Fourier-transform infrared (FTIR) spectra for composite insulators is collected to establish an aging classification dataset. A deep learning model based on a one-dimensional convolutional neural network (1D-CNN), termed FTIR Aging Net (FAN), is proposed. This model effectively leverages the entire spectral range of FTIR to accurately classify the aging state of composite insulators, achieving a classification accuracy of 98.44%. This high level of accuracy underscores the model’s ability to capture discriminative features from the FTIR for reliable aging assessment. An interpretability analysis of the model reveals that, during the early stages of aging, the primary degradation occurs in the Si-O backbone and C=C bonds, while in the later stages, further breakage of C-H bonds is observed. FAN facilitates the assessment of composite insulator aging, providing valuable insights into the aging mechanisms of the material. This can serve as a technical support tool for the routine maintenance and operation of composite insulators.

Silicone rubber of composite insulators is extensively utilised because of its superior hydrophobicity and hydrophobicity transfer characteristics, especially in heavily polluted environments. This work investigates the hydrophobicity transfer process of silicone rubber materials with different insoluble substances and explores the impact of microscopic parameters on hydrophobicity transfer. Experiments employed diatomaceous earth and kaolin of various particle sizes and origins as inert substances to analyse their effects on hydrophobicity transfer characteristics. Results indicate that diatomaceous earth, with its porous structure, facilitates faster transfer compared to the dense structure of kaolin. Pore size and specific surface area are critical parameters influencing transfer rates. Larger outer pores accelerate the initial ‘fast process’ while smaller inner pores govern the subsequent ‘slow process’ Fractal characteristics of pores also affect transfer efficiency with higher fractal dimensions leading to more extensive transfer. The hydrophobicity transfer process in silicone rubber involves dynamic diffusion influenced by the complexity of internal channels and surface structures. Enhanced understanding of these mechanisms can improve the performance and reliability of composite insulators in polluted environments.