Stochastic wave spectra estimation (SWSE) based on response surface methodology considering uncertainty in transfer functions of a ship

Estimating waves via measured ship responses

“…In this context, it should be noted that how accurate the wave height, for instance, needs to be, depends on the use-case; in some cases a low accuracy is fine, while in other cases we would like high accuracy. On the same note, it is also interesting to mention that a physics-based version of the WBA, making it possible to include the uncertainty in the transfer functions of the ship, has been developed 26 . In the future, it appears interesting to investigate possible combinations of that version with the hybrid framework discussed in the present paper.…”

“…In contrast thereto, purely data-driven approaches exist, relying on machine learning. The main advantage of physics-based methods is that they can provide estimates of the complete directional wave spectrum, where different mathematical strategies 13 – 26 are applied for solving the highly underdetermined equation system describing the physical relationship between the energy densities of waves and the corresponding responses. In machine learning methods 27 – 36 , the general relationship between wave parameters and induced responses of a particular ship is learned by comprehensive training with large datasets of measured responses against available sea state information, e.g.…”