Based On The Neural Network Modeling And Prediction Of Ship Attitude Motion

There are a large of nonlinear systems in practice and ship movement system is one of them. Ships can have 6-motions complex movements because of the ocean waves, the sea breeze and other disturbances, which have the very randomness and the non-linearity. So it is very important to do non-linear ship modeling and prediction for ship's safe sailing.Neural network is one of the main technologies of intelligent control currently. It simulates the structure, the memory and the information processing of the human brain, with a good character of studying useful knowledge from input and output data. This paper leaves from the neural network theory, fully studied back propagation network and radial basis function network and their improved algorithm. Then use them to do ship modeling and prediction and get good results.Firstly, collect and collate the correlation resources about the ship motion posture prediction technology in domestic and foreign. Study the elementary theory on neural network and research to the back propagation neural network thoroughly, and then use the improved conjugate gradient algorithm to the BP neural network, therefore it is enabled to improve the network performance and increase the prediction precision.Secondly, research radial basis function network and its improvement—generalized regression neural network, and use them to model and predict the ship motion, and reach good results. Analyze and compare the prediction precision and the time of training between those algorithms.Finally, discuss and research how to select the number of the neural units and the training data.