Wind Farm Drone Edge-Assisted Diagnosis System: Lightweight Deep Learning Implementation

With the rapid development of the wind energy industry, efficient operation and maintenance (O&M) of wind farm equipment is crucial. This paper proposes a lightweight deep learning-based UAV edge diagnostic system for wind farms, aiming to achieve rapid and accurate fault detection in wind turbines. By adopting a lightweight convolutional neural network (CNN) model, the computational complexity is reduced, enabling efficient execution on UAV edge devices. Experiments conducted on a real-world wind farm dataset demonstrate that the system maintains high diagnostic accuracy (reaching 92.3%) while significantly reducing the average diagnosis time (from 12 seconds with traditional methods to 2.5 seconds). This effectively enhances wind farm O&M efficiency and provides reliable technical support for intelligent O&M in wind farms.