In order to control complicate nonlinear system accurately, it is very important to acquire mathematical model of the nonlinear system. Nonlinear system identification is one of the several ways of the acquisition of mathematical model of the nonlinear system. Therefore, the research on the identification of nonlinear system is very important. The research object of the paper is an autonomous underwater vehicle developed by HEU underwater vehicle lab. The AUV is an nonlinear coupling system. In order to acquire the model of the AUV, two identification methods is presented and discussed.First, the linear and nonlinear maneuverability equations of the Underwater Vehicle (UV) are established on the basis of hydrodynamic analysis, and then transformed to discrete equations. In the need of UV's mathematical model and the restriction of physical conditions, the Zigzag test has been designed in this paper and performed in the sea trial to obtain practical experimental data. With the data, the hydrodynamic coefficients of the UV have been estimated by SI methods including the Neural Network (NN) method and the Kalman Filter (KF). In the end, in order to verify the reliability of the model, the motion simulation system is established, and the real-time motion simulation is carried out by solving the equations of motion. Briefly, this paper obtains hydrodynamic coefficients of the UV, and the motion simulation shows the model is reliable. The technique of SI is practicable for the research of UV s maneuverability and adaptive control. |