Frequency-based model validation and parameter identification of a sea-surface vehicle

This paper focuses on the modeling of the roll motion for a manned sea-surface vehicle. An extension of the Conolly model is proposed, where the uncertainties in the mechanical torques acting on the vehicle – torques coming out from wind and sea-wave action – are also included in a general state space representation. A frequency-based identification procedure is provided in order to identify both the ship and the disturbance parameters. The algorithm has been tested on real data and most significative results are proposed.