A Novel Wind Speed Estimation Based on the Integration of an Artificial Neural Network and a Particle Filter Using...

Area of Science:

• Oceanography
• Remote Sensing
• Geophysics

Background:

• Ocean wind speed is critical for climate change research and maritime safety.
• Global Navigation Satellite Systems - Reflectometry (GNSS-R) offers a promising method for ocean observation.
• Accurate wind speed estimation is essential for understanding ocean dynamics.

Purpose of the Study:

• To develop a novel technique for precise ocean wind speed estimation.
• To integrate artificial neural networks with particle filters for improved accuracy.
• To validate the proposed method using real-world data, including typhoon events.

Main Methods:

• Utilized Global Navigation Satellite Systems - Reflectometry (GNSS-R) for signal reflection analysis.
• Integrated an artificial neural network with a particle filter, optimized by particle swarm optimization.
• Employed BeiDou Geostationary Earth Orbit (GEO) satellite data and in situ buoy measurements for validation.

Main Results:

• The proposed technique demonstrated effective wind speed estimation.
• Achieved a root mean square error of approximately 1.9 m/s in wind speed measurements.
• Successfully validated results against in situ data during typhoon events.

Conclusions:

• The novel GNSS-R based technique shows high potential for accurate ocean wind speed monitoring.
• Artificial neural network and particle filter integration enhances estimation precision.
• This method contributes to improved climate change studies and maritime management.