Feasibility Study for Predicting Collision Possibility Sea Area for Each Ship by Using Support Vector Machine
Feasibility Study for Predicting Collision Possibility Sea Area for Each Ship by Using Support Vector Machine
复制标题
DOI：
10.1109/smc.2018.00143
复制
发表时间：
2018-10
期刊：
影响因子：
--
通讯作者：
T. Okazaki;M. Terayama;Chihiro Nishizaki
中科院分区：
文献类型：
--
作者：
T. Okazaki;M. Terayama;Chihiro Nishizaki
本文提出了一种新的方法来支持船舶在湾区避让航行，通过网络系统向每艘船舶指示碰撞概率海域。这项研究集中在海上航线出口处的碰撞。然后，通过学习海上交通数据的支持向量机对出口点的船舶新航向进行预测。在考虑障碍区的情况下，利用预测的船舶新航向数据，计算了每艘船出口位置的碰撞概率海域面积。仿真结果验证了该方法的有效性。
This paper proposes a new method to support the evading navigation of a ship within a bay-area by indicating the collision probability sea area to each ship through a web system. This study focused on collision at the exit of a sea route. Then, the ship's new course at the exit point was predicted by the support vector machine, which studied marine traffic data. The collision probability sea area at the exit position was calculated for each ship by considering the obstacle zone through the proposed approach, and the data of the predicted ship's new course. The effectiveness of the proposed approach was demonstrated by the simulation results.