Pressure derived wave height using artificial neural networks

Underwater ultrasonic acoustic transducers are widely used for ocean wave measurements, since they measure surface wave directly. However, their effectiveness may be severely affected under rough sea conditions. In which breaking waves generate bubbles, which in turn interfere with acoustic signals. Therefore, when the seas are rough, one often has to rely on pressure transducer, which is generally used as a back up for the acoustic wave gauge. Then one uses a pressure transfer function to obtain the surface wave information. This study used the artificial neural network to convert pressure signal to significant and maximum wave height, using data obtained from various water depths. The results showed that the wave height obtained from the artificial neural network was more accurate than that from using linear pressure transfer function for water depth larger than 20 m.