Modeling of the nonlinear drift oscillations of moored vessels subject to non-Gaussian random sea-wave excitation

A quadratic system model based on Volterra series representation is utilized to model the nonlinear response of moored vessels subjected to random seas. The key idea is to represent the relationship between the incident sea wave (input) and corresponding sway response of the moored vessel (output) with a parallel combination of linear and quadratic transfer functions, and to estimate them by processing actual input and output data. Compared to previous approaches, we take the important step of removing the restriction that the random input must possess Gaussian statistics. The feasibility and validity of the approach is demonstrated by analyzing experimental data taken in model basin tests. We also describe some of the deleterious consequences of assuming Gaussian sea-wave excitation when in fact the excitation is non-Gaussian.