Real-Time Wind Sensing With Massively Parallel Sample Processing

Light detection and ranging (LiDAR)-based wind monitoring has significant implications in various fields, including aviation and wind energy. Accurate and real-time wind speed measurement is crucial for the safety and the performance of these applications. This letter presents an field programmable gate arrays-based real-time wind LiDAR system. The proposed solution incorporates a massively parallel sample processing architecture that uses 16 parallel fast Fourier transform units and 256 unique spatial measurement ranges. A real-time noise suppression averaging technique is implemented to further reduce processing overhead. Field experiments highlight the superiority of the presented solution against state-of-the-art wind LiDARs, as it achieves better error floor (5%), better spatial resolution (49%), and very low processing latency (120 ms).