Identification of Motion Model Parameters for a Surface Ship under Disturbances

Abstract

An approach to identifying the parameters of the motion model for a surface ship subject to external disturbances is proposed. It uses the measurements of heading, yaw rate, speed through the water, and global satellite navigation system (GNSS) data. The model structure is set in the state space. We use the criterion of how close is the real vehicle response to a given control input to its motion model response under the same disturbances. It is proposed to apply the Kalman filter with the state vector including the disturbances, and an iterated procedure for estimating parameters by minimizing the criterion. It is shown that this ensures stable identification of model parameters under different disturbances. Simulation results are presented to evaluate the quality of identification. The approach was validated in full-scale tests of a high-speed boat.