Marine vehicle modeling and tracking using wavelet type nets model

The main objective of this paper is to address the tracking control problem of a ship moving with a constant velocity along a desired path. To this end, a combination of mathematical model (MM) and statistical model, namely the expanded exponential weight wavelet network based ARX (E-EW-WNN-ARX) model is proposed as internal predictor of a model predictive controller (MPC). Firstly, the EW-WNN-ARX model is used to describe the ship motion between the difference of heading angle deviation and the rudder angle of the ship. To represent the ship's motion nonlinearity, the rolling angle is used as the EW-WNN-ARX model index to allow the model parameters to vary with the ship moving state. The EW-WNN-ARX model is identified off-line by using previously observed real data. Then The difference of heading angle deviation are expanded and integrated into mathematical model (MM) characterizing the position tracking error of the ship to form the E-EW-WNN-ARX-MM model which is used to develop a state space model that represent the ship tracking motion behavior. Finally, the identified ship's state-space type tracking motion model is used as predictor of a model predictive controller to steer ship moving forward with constant velocity along a predefined reference path. The effectiveness of the proposed modeling and control methods are demonstrated by the tracking control simulation in which the modeling data was obtained from the Shioji-maru experimental ship of Tokyo University of Marine Science and Technology of Japan.