Research on Intelligent Assessment of Insulators Based on GATr-FT

With the rapid development of intelligent inspection for transmission lines, traditional insulator condition assessment methods have difficulty meeting recognition requirements in complex scenarios. To improve the accuracy of intelligent insulator assessment, this paper proposes an intelligent insulator assessment method based on GATr-FT. First, insulator images are constructed as graph-structured representations to enhance the modeling capability of spatial topological relationships. Then, GATr is employed to extract condition features containing both local details and global geometric information. Finally, a fine-tuning strategy and a multi-task output mechanism are introduced to achieve insulator defect recognition and condition assessment. Experimental results show that the proposed method achieves good performance in terms of assessment accuracy, generalization capability, and robustness, providing technical support for intelligent inspection of transmission lines.