Identification Of Mathematic Model Of The Model Ship

Maneuverability is one of the important properties of the ship, which determine the economy and security of the navigation to some extent. According to the ship maneuverability standards developed by the International Maritime Organization, ship manoeuvrability should be predicted at the ship design stage. At present, computer simulation based on mathematical model is a practical and effective method for ship maneuverability prediction and evaluation in the ship design stage. To use this method, accurately determining the mathematical model is vital. What’s even more critical is how to determine the parameters in the mathematical model. Over these years, System identification techniques has made remarkable progress in ship maneuverability, which fully demonstrates that it is a fairly effective method.The main work of this paper is to work out a ship modeling in the laboratory. We get the input and output data from the zig-zag test, and filter the test data by using the wavelet denoising technique. The study uses several classical system identification methods, including least squares method, extended Kalman filter, extended least squares method. In addition, it improves the method of extended least squares and applies extended matrix to the gradient correction algorithm. Thus a hybrid identification algorithm which combining extended least squares method with gradient correction method is proposed. Combined the experimental data with the identification algorithm mentioned above, I identified the second-order and third-order response model of the model ship respectively. Ultimately, parameters in ship maneuvering equations of motion are confirmed and predicted.The mathematical model is an approximate description of the actual system. It is impossible to make an accurate simulation of the actual situation. What we need to find is only the optimal model in the model class. After analysis of the prediction and experimental results for each model, the effectiveness and superiority of this method is proved effectively.