Hilbert transform-based nonparametric identification of nonlinear ship roll motion under free-roll and irregular wave exciting conditions

This paper presents a numerical investigation of two Hilbert transform (HT)-based nonlinear identification methods to estimate dynamic characteristics of ship roll motions using free-roll response. Since the methods are nonparametric, they can avoid the identification error caused by false model assumption. Then, these methods are extended to identify roll motion from random ship response in irregular waves. A novel two-step identification scheme is proposed by combining the random decrement technique and the HT-based identification methods. Results of numerical studies show that the two methods are effective for roll motion identification using free-roll response. More precise results for the nonlinear coefficients can be obtained when multiple harmonics were considered. Four wave exciting conditions with different significant heights and peak periods were simulated to investigate the proposed two-step scheme. The results indicate that the nonlinear dynamic characteristics of roll motions can be successfully identified from random ship vibration under resonant conditions.