摘要
内河航行环境中AIS信号拥塞和丢失现象严重,导致船舶AIS轨迹稀疏不连续,而传统的采用数值分析方法拟合船舶丢失运动轨迹,无法适应地形弯曲复杂的内河航道水域,容易出现穿越陆岸的问题.为此,本文提出一种基于航行经验的内河船舶轨迹估计方法,针对稀疏轨迹的相邻采样点,基于空间一致性、时间一致性和航向一致性,提取有效的他船航行轨迹,通过核密度估计方法建模航行经验—航行热度分布;然后,考虑船舶运动趋势,计算得到基于航行热度的最优路径,并将其作为近似估计轨迹.选取三种典型内河航道场景进行稀疏轨迹还原实验.结果表明,本文方法能够很好地适应内河航道的复杂地形.
In the inland navigation environment,the serious AIS signal congestion and loss result in the ship AIS trajectory to be sparse and discontinuous. Traditional numerical analysis method is used to fit the lost trajectory of ships,and it cannot adapt to the complicated inland waterways with complicated bend,and is prone to cross the land coast. In this paper,a trajectory estimation method for inland river ship was proposed based on navigation experience. Based on the spatial consistency,time consistency and heading consistency,the history similar ship trajectories were extracted for neighbor sampling points of sparse trajectory,and the navigation experience was estimated by the method of Kernel density estimation,and navigational heat distribution map was modelled. And then the optimal route was calculated as approximate estimation trajectory in consideration of ship movement trend based on navigation heat. Three typical inland river channel scenarios were selected to carry out sparse trajectory restoration experiments.Results show that the proposed method can adapt well to the complex terrain of the inland waterway.
出处
《大连海事大学学报》
CAS
CSCD
北大核心
2017年第3期7-13,共7页
Journal of Dalian Maritime University
基金
国家自然科学基金资助项目(51679180)
武汉理工大学双一流项目资助
关键词
内河船舶轨迹
自动识别系统(AIS)
航行经验
核密度估计
轨迹估计
trajectory in inland waterway
automatic identification system（AIS）
navigation experience
Kernel density estimation
trajectory restoring
作者简介
黄亮（1986-），男，博士后，E-mml：plaquemine@whu．edu．cn
文元桥（1975-），男，博士，教授，博士生导师，E-mail：wenyqwhut@foxmail．com．
