摘要
为避免船舶发生碰撞,提高船舶航行安全性,本文首先建立船舶碰撞模型。基于有限样本点,采用神经网络算法对其进行聚类分析,使得样本拟合误差收敛,具有较高的稳定性,并且收敛精度较高,可为工程实践提供参考。
In order to avoid ship collision,and improve the safety of navigation of ships,this paper established the model of collision first. Based on limited sample points,using the neural network algorithm for clustering analysis,makes the sample fitting error convergence and has high stability,this method can provide reference for engineering practice.
出处
《舰船科学技术》
北大核心
2016年第14期163-165,共3页
Ship Science and Technology
关键词
碰撞
聚类分析
神经网络算法
F分布
collision
clustering analysis
neural network algorithm
F distribution
