Abstract
Large amplitude rolling in random beam seas is studied by moment equations and stochastic excitation is represented through a cascading filter driven by pure Gaussian white noise in order to apply Markov approximation. The moment equation is generated from a six state space rolling model using the Itô differential rule. The Gaussian and non-Gaussian cumulant neglect method is applied to close the infinite hierarchy of moment equations. In this paper, an automatic neglect tool is developed to handle the complex and untraceable higher-order cumulant neglect closure method to capture the non-Gaussian effect of nonlinear rolling phenomena. Both transient and stationary responses of statistical moments are obtained through the solution of the closed moment equations.