Classification and Diagnosis of Propeller Damages Using FFT-Based Motor Current Signature Analysis of DC-Link Current

This study proposes a frequency-domain diagnostic approach for detecting and characterizing unmanned aerial vehicle(UAV) propeller damage using only motor electrical signals. Motor current signals were analyzed via fast Fourier transform (FFT) across 36 systematically defined damage scenarios covering multiple locations and severity levels. The results demonstrate that specific harmonic components of the DC-link current exhibit strong and consistent sensitivity to propeller damage, with the sixth harmonic showing the most pronounced deviation from the normal state. Quantitative analysis revealed clear location-dependent trends: leading-edge damage produced frequency standard deviations up to 93.5 Hz, approximately 2.26 times higher than trailing-edge damage, while tip damage caused progressive bandwidth expansion, reaching 322.07 Hz, the largest variability observed. In contrast, hub-proximal damage exhibited minimal deviation, with bandwidths as low as 108.7 Hz. These findings confirm that motor current signals alone enable reliable discrimination of propeller damage presence, type, and coarse location, providing a practical and scalable solution for sensor less UAV ground inspection and maintenance.