ニューラルネットワークを用いた推進性能の推定

－バラスト状態への適用－

抄録

The neural net work has been actively used in the various fields of technology as an effective tool of non-linear approximation technique. In the previous paper, authors showed that the neural net work could estimate propulsive performance in good accuracy for full load condition of some fine ships. In the present study, author shows the results of the relation between the ship form parameters and propulsive performance investigated with the neural net work by use of 3200 cases of model test results collected from the existing data base. As the results it was shown that propulsive performance of various type of single screw displacement ships in both full load and ballast condition with trim can be estimated in good accuracy.