Online modeling and prediction of maritime autonomous surface ship maneuvering motion under ocean waves

Abstract

The development of maritime autonomous surface ships (MASS) has attracted much attention at present. As a necessary technology of MASS, ship motion prediction still has some problems, such as low efficiency, poor real-time performance, and easy to be affected by sea conditions. Aiming at the above problems, based on the mathematical model of ship maneuvering motion, this paper uses the least squares support vector machine (LS-SVM) identification algorithm, to establish an offline black-box identification model of MASS motion considering rudder angle and propeller revolution rates, and effectively predicts the motion of ships. On this basis, a sliding window is designed, and an error-based online modeling method of ship motion is proposed to realize adaptive prediction of MASS motion under the effect of ocean waves. A large MASS was selected for modeling and simulation experiments. The results show that compared with the traditional LS-SVM modeling algorithm, the model built by this method has a strong adaptive ability and can maintain high accuracy for a long time, meeting the needs of real-time prediction of MASS motion in practical applications.