Loading…
NECEC 2017 has ended
View analytic
Wednesday, November 15 • 10:00 - 10:20
The Use of Artificial Neural Networks in Hindcasting and Filling Gaps in Buoy Wind Speed Data Under Extreme Winds

Sign up or log in to save this to your schedule and see who's attending!

The present work approaches the use of artificial neural networks in hindcasting and filling gaps in buoy wind speed data during events of extreme wind. The chosen network architecture is a nonlinear auto-regressive neural network with exogenous entries. Significant wave height was chosen as the input after cross-correlation analysis, which determined the number of input and feedback delays. In order to test the method, a data set from a buoy in Placentia Bay, NL during the extreme wind event of March 11, 2017 was used, in comparison to other wind estimation methods, i.e. the Sverdrup-Munk-Bretschneider (SMB) relationship, power series regression, and backpropagation neural networks. The presented method outperformed all other techniques, with a mean absolute error below 1 m/s and correlation coefficient of 0.95 during hindcasting, and it was able to fill the gaps in the data following the trend of other weather stations positioned close to the buoy, proving the efficacy of the method.


Wednesday November 15, 2017 10:00 - 10:20
Salon C 180 Portugal Cove Rd, St. John's, NL A1B 2N4, Canada