| 基于隐马尔科夫模型的VLCC目的港预测 |
| |
| 引用本文： | 杨春,王一丹,徐晖,胡勤友,潘亚兰.基于隐马尔科夫模型的VLCC目的港预测[J].上海海事大学学报,2020,41(4):42-49. |
| |
| 作者姓名： | 杨春 王一丹 徐晖 胡勤友 潘亚兰 |
| |
| 作者单位： | 上海海事大学商船学院,上海201306;招商局能源运输股份有限公司,广东深圳518067 |
| |
| 基金项目： | 上海市科学技术委员会社会发展领域重大项目（18DZ1206300） |
| |
| 摘 要： | 超大型油船(very large crude carrier，VLCC)目的港预测对海运原油流向预测以及货源地未来运力估计具有重要作用。针对VLCC的AIS目的港信息存在缺失、更新不及时、不准确等现象，提出一种基于隐马尔科夫模型的VLCC目的港预测方法。分析船舶AIS轨迹数据，得到油船历史停靠港口序列；根据VLCC轨迹提取习惯航路，以航路中的交叉点为依据设置观测线；利用船舶航行轨迹数据判断船舶是否经过观测线以及经过观测线的方向，对不同方向分别计算船舶在挂靠港间的转移概率矩阵和船舶挂靠港与观测线间的输出概率矩阵，建立VLCC目的港预测模型并进行预测。研究结果表明：在大多数情况下VLCC目的港预测的准确率可以达到70%以上；航线越固定、运行越规律的船舶，预测准确率越高；船舶越靠近目的港，预测越准确；重载状态下的船舶目的港预测更准确。
|
| 关 键 词： | 智能交通 AIS轨迹数据 超大型油船(VLCC) 目的港预测 隐马尔科夫模型 |
| 收稿时间： | 2020/3/17 0:00:00 |
| 修稿时间： | 2020/6/5 0:00:00 |
VLCC destination port prediction based on hidden Markov model |
| |
| Institution: | Shanghai Maritime School,Shanghai Maritime School and Shanghai Maritime School |
| |
| Abstract: | Destination port prediction of very large crude carriers (VLCCs) plays an important role in forecasting the flow of maritime crude oil and estimating the future capacity of the source. Aiming at the fact that VLCC destination port information from AIS is missing, untimely and inaccurate, a prediction method for VLCC destination ports is proposed based on the hidden Markov model. The AIS trajectory data are analyzed to obtain the historical calling port sequence. The customary route is extracted based on the VLCC trajectory, and the route observation line is set based on the route intersections. The ship trajectory data are used to determine whether the ship passes through the observation line and the direction through the observation line, the transition probability matrix of the ship between the calling ports and the output probability matrix between the ship calling port and the observation line in different directions are calculated, and a prediction model for VLCC destination ports is built to carry out the prediction. The research results show that: in most cases, the accuracy of the VLCC destination port prediction can reach more than 70%; the more fixed the ship route and the more regular the ship operation, the higher the prediction accuracy; the closer the ship is to the destination port, the more accurate the prediction is; the ship destination port prediction under heavy load is more accurate. |
| |
| Keywords: | smart transportation AIS trajectory data very large crude carrier (VLCC) destination port prediction hidden Markov model |
| 本文献已被 万方数据 等数据库收录！ |
| 点击此处可从《上海海事大学学报》浏览原始摘要信息 |
| 点击此处可从《上海海事大学学报》下载免费的PDF全文 |
