Anomaly Detection Method for Substation Equipment Based on Feature Matching and Multi-Semantic Classification

Abstract

The health status of substation equipment is an important guarantee for the safe and stable operation of substation, and promptly detection of anomaly conditions in substation equipment can avoid serious safety hazards. To address this issue, this paper proposes an anomaly detection method for substation equipment based on image registration and deep learning. First, the anomaly categories of substation equipment are detected based on multi-semantic feature network and fine-grained classification. Then, a sparse cross-domain feature matching algorithm is introduced to register the image of substation equipment, and edge detection and image denoising is used to detect the anomaly areas in the image. Finally, the image registration and multi-semantic recognition are merged for the integrated anomaly detection. The experimental results illustrate that the proposed method can rapidly and accurately detect the anomaly of substation equipment, and significantly improve the automation level of substation equipment and the safety of substation operation.