A prediction method of ship maneuverability based on Least Square Support Vector Machine (LS-SVM) is proposed, which has excellent nonlinearity approximation ability and better generalization ca-pability. The nonlinear model is established by the outputs about maneuverability corresponding to the in-puts concerned to principal particulars of the ship. The prediction model is applied to the turning motion of free-running model test. Comparing the test data and predicted results, the predicted results are in good agreement with the test data. The test cases illustrate that the proposed method is very successfully in fore-casting the ship maneuverability.% 文章提出了基于LS-SVM的船舶操纵性黑箱建模方法,通过选取影响操纵性能的主要因素,利用LS-SVM建立了操纵性能的非线性回归模型,并利用回归模型对自航模回转试验进行了预报,通过试验与预报结果的对比分析表明,所建立的回归模型具有较高的预报精度,为船舶操纵性预报提供了一种新的技术手段。
展开▼