A hybrid model framework which assimilates measured motion data with physics-based forecasts of the heave response of a semisubmersible using Bayesian statistical methods is demonstrated. The hybrid model treats for errors inherent in the physics-based forecasts due to misspecification of the numerical wave spectra and response amplitude operator (RAO) inputs, enabling bias corrected probabilistic heave forecasts. Model diagnoses motivate an additional level of complexity required for the error structure in the Bayesian model, specifically to account for heteroskedasticity and time-correlated errors. The hybrid model forecasts have been evaluated during periods where the heave resonance and cancellation frequencies were excited. The method is demonstrated to be effective for providing reliable quantification of uncertainty and allows for sequential model updating. This justifies its value for improving the efficiency and safety of offshore operations.
Milne et al. (Thu,) studied this question.