Uncertainty analysis of the hydrodynamic coefficients estimation of a nonlinear manoeuvring model based on planar motion mechanism tests

Abstract

Uncertainty analysis of the identified hydrodynamic coefficients of a nonlinear manoeuvring model is presented in this paper. The classical parameter estimation method, Least Square, is briefly introduced, and the uncertainty of the hydrodynamic coefficients due to the noise in the measured data is analysed using singular value decomposition. Then, two methods, truncated singular values decomposition and Tikhonov regularization, are introduced to diminish the uncertainty. A nonlinear manoeuvring mathematical model of a marine surface ship is derived using Lagrange's method. The dimensionless hydrodynamic coefficients are obtained using the Least Squares method, truncated singular values decomposition and Tikhonov regularization with Planar Motion Mechanism test data. The validation process is carried out to test the performance and accuracy of the resulting nonlinear manoeuvring models. The result shows that identification of the uncertain parameters using the truncated singular values decomposition and Tikhonov regularization resulted in good estimating the parameters and significantly diminish the uncertainty.