Nonparametric Identification of Roll Damping and Nonlinear Restoring Forces for a Ship from a Free Roll Decay Simulation

摘要

Roll damping and nonlinear restoring forces have an important influence on ship roll motion. To identify them from free roll decay behaviors, the models are generally assumed to be known a priori. Here, the stochastic inverse method is introduced, and a new nonparametric approach is adopted to identify roll damping and nonlinear restoring forces from a free roll decay simulation, which is carried out by the Reynolds-averaged Navier-Stokes method and overset mesh technology. In the nonparametric identification of roll damping and nonlinear restoring forces, the nonlinear Volterra integral equation of the first kind is solved by the Markov chain Monte Carlo (MCMC) method. The conditional expectation and covariance of the conditional component distribution used in the Gibbs sampler are derived directly in the matrix-vector form by using an analytic inversion formula. Reconstructed roll histories are compared with ones from both the earlier work of the authors and the computational fluid dynamics method. Results show that the MCMC method is an effective way to identify roll damping and nonlinear restoring forces from free roll decay.