Transmission Line Fault Detection Method Based on Multi Source Data Fusion and Edge Computing

Aiming at the complex operation environment of transmission lines and the poor recognition effect of small target defects, a transmission line fault detection method based on multi-source data fusion and edge computing is proposed. Firstly, an edge computing controller is designed to preprocess the data at the edge of the system. Then, based on orthogonal basis forward neural networks, multi-source data fusion is achieved. Finally, the YOLOv7 algorithm is improved by introducing embedded attention modules, shallow weighted feature fusion networks, and Focal-CIOU loss functions, which will be used to process fused multidimensional data and output fault detection results for transmission lines. Based on the Pytorch platform, experimental tests were conducted on the proposed method, and the results showed that its average detection accuracy and detection speed were 90.26% and 39 fra<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$m$</tex>es/s, respectively, which can meet the needs of intelligent development of transmission lines.