Abstract

A feasibility study of Bayesian wave estimation was carried out to investigate the influence of ship response functions to the estimated wave parameters. The response functions are used to estimate the directional wave spectra from the measured ship motion data. That is to solve an inverse problem of how to estimate the input from the measured output data. It is theoretically favorable to use the correct response functions but the ship course and speed measured by GPS are not match well with actual ones because of the current speed and so on. The mismatch causes to chose incorrecct response functions and large estimation error consequently will occur. In this study, a practical method for modifying the response functions is proposed. Using the modified response functions, the wave parameters are estimated without large estimation error. In order to examine the proposed concept, the estimated wave parameters were compared with observed ones. Furthermore, ABIC surfaces with two hyperparameters were also investigated between the estimates based on actual and modified response functions. It was confirmed that the corrected response functions can be used for Bayesian wave estimation without large estimation error.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call