An Efficient Method for Predicting Long Term Extreme Design Forces of Wave Energy Converters

Yingguang Wang
DOI: https://doi.org/10.1016/j.apor.2022.103094
IF: 3.761
2022-01-01
Applied Ocean Research
Abstract:This paper proposes to utilize a modified adaptive KDE (Kernel Density Estimation) method based on linear diffusion process in order to more efficiently and accurately calculate the sea state parameter distribution tails and to extrapolate well. This modified adaptive KDE method is applied in predicting the probability distribution tails of a measured ocean wave dataset at National Data Buoy Center Station 44,009, and its efficiency and accuracy have been clearly substantiated through comparisons with the prediction results via the parametric method and the traditional adaptive KDE method. Subsequently, the modified adaptive KDE method is utilized for deriving an accurate 50-year environmental contour line based on the aforementioned measured wave dataset. The derived environmental contour line and another contour line obtained using a parametric contour approach are then applied for predicting the 50-year design PTO (Power-Take-Off) force values for a point absorber Wave Energy Converter (WEC). The predicted 50-year design PTO force value based on the proposed modified adaptive KDE contour (whose accuracy has been clearly substantiated in this paper) is 20674KN, and the WEC PTO system designed based on this accurate prediction obviously will be safer and more reliable. On the contrary, the forecasted 50-year design PTO force value based on the inaccurate contour via the parametric method is only 9256.3 KN, which is likely to cause failure of the PTO system and the entire Wave Energy Converter.
What problem does this paper attempt to address?