Image feature-based aircraft wake identification with coherent Doppler wind lidar

Area of Science:

• Aerospace Engineering
• Computer Vision
• Fluid Dynamics

Background:

• Aircraft wake vortices pose significant risks to aviation safety.
• Vortex identification is crucial for efficient airport operations.
• Existing methods may lack speed or require extensive datasets.

Purpose of the Study:

• To develop a lightweight and accurate wake vortex identification method.
• To enable rapid localization of vortex core regions.
• To provide a viable solution for real-time monitoring systems.

Main Methods:

• Utilizes the Hue, Saturation, and Value (HSV) color space and image texture features.
• Transforms lidar wind field data into HSV color space to leverage velocity gradients.
• Relies on inherent data properties, not pre-collected datasets or preset parameters.

Main Results:

• Achieved root mean square errors (RMSE) of 2.57m and 6.08m for vortex core positions in simulations.
• Circulation retrieval RMSE of 38.14m²/s (10.37%) validated against numerical simulations.
• Successfully captured wake evolution in field experiments at Guangzhou Baiyun International Airport.

Conclusions:

• The proposed method offers accurate and rapid wake vortex identification.
• It significantly reduces computational complexity for real-time applications.
• Demonstrates effectiveness in both simulated and real-world airport environments.