Research on the Algorithm Model for Measuring Ocean Waves Based on Satellite GPS Signals in China

Area of Science:

• Oceanography and Marine Technology
• Signal Processing and Data Analysis
• Artificial Intelligence in Environmental Monitoring

Background:

• Global Positioning System (GPS) wave buoys offer in situ wave monitoring using satellite signals.
• Traditional GPS wave measurement relies on Doppler frequency shifts to calculate buoy velocity and wave parameters.
• A key limitation is data anomaly due to intermittent GPS signal loss when the antenna is submerged.

Purpose of the Study:

• To improve the accuracy and reliability of GPS wave buoy measurements.
• To address data anomalies caused by signal loss without extending data acquisition time.
• To validate the performance of a novel GPS wave buoy against established technologies.

Main Methods:

• Implemented a neural network algorithm to correct anomalous buoy velocity data.
• Applied a Gaussian low-pass filter to refine the raw directional wave spectrum.
• Conducted comparative on-site sea tests using the SBF7-1A GPS wave buoy and a DWR-MKIII Waverider buoy.

Main Results:

• The neural network effectively corrected unusual velocity data, negating the need for extended signal acquisition and saving battery power.
• Gaussian filtering reduced noise in the directional wave spectrum, enhancing measurement precision.
• Ocean wave parameters (wave height, period, direction, spectra) from the SBF7-1A GPS buoy showed good consistency with the Waverider buoy.

Conclusions:

• The proposed improvements, including neural network correction and Gaussian filtering, significantly enhance GPS wave buoy performance.
• The SBF7-1A GPS wave buoy demonstrates accuracy comparable to traditional gravity-acceleration-type wave buoys.
• The study confirms the feasibility and validity of the advanced techniques for reliable ocean wave monitoring.