文摘
|
随着港口信息化建设的推进,积累了大量来源多样、结构各异的海事大数据,为了解港口城市的生产力和区域经济发展水平提供了新的契机。本文综合介绍了作者近期关于如何利用海事大数据进行港口感知计算的工作,给出了一个基于海事大数据的港口感知计算框架,利用船舶GPS轨迹、船舶属性、港口地理信息和港口设施参数等多源异构海事大数据,估算出一系列反映港口生产力的指标,从而对港口进行综合评价和比较。首先,利用船舶轨迹和港口地理信息数据,自动检测船舶在港口码头中的靠泊装卸事件;然后,利用船舶属性和港口设施数据,自动估计出每次靠泊装卸事件的货物吞吐量;最后,对各个港口码头的靠泊船数和货物吞吐量进行统计,从而计算出一系列港口生产力指标,包括到港船数、货物吞吐量、码头作业效率和泊位利用率等。在2011年的海事大数据上的实验结果表明,本框架能准确地估算出上述港口生产力指标。同时,以香港为例对上述港口的生产力指标进行分析,探讨基于海事大数据的港口感知计算框架在提高港口生产效率、优化海运航线中的积极作用。 |
其他语种文摘
|
With the wide applications of information and communication technologies in port infrastructures and operations, huge volumes of maritime sensing data have been generated. These data come from various sources and demonstrate heterogeneous structures, providing us with new opportunities to understand port performance and regional economic development. In this paper, we introduce the recent work on port sensing and computation based on maritime big data. Specifically, by making use of ship GPS trajectories, ship attributes, port geographic information and port facility parameters, we can automatically estimate a set of metrics for the measurement and comparison of port performance. First, we can use ship GPS trajectories and port geographic information to detect the events of ships arriving at different ports and terminals. Second, we can use ship attributes and port facility parameters to estimate the cargo throughput of each arrived ship. Third, we can aggregate the ship arriving events and the cargo throughput in different terminals and ports to derive a set of port performance metrics, including ship traffic, port throughput, terminal productivity and facility utilization rate. Evaluation results using real-world maritime data collected in 2011. Results showed that these methods accurately estimated the port performance metrics. We also presented a case study in port of Hong Kong to showcase the effectiveness of our framework in port performance analysis. |
来源
|
地球信息科学学报
,2016,18(11):1485-1493 【核心库】
|
DOI
|
10.3724/SP.J.1047.2016.01485
|
关键词
|
海事大数据
;
港口
;
城市感知
;
城市计算
;
数据挖掘
|
地址
|
1.
浙江大学计算机学院, 杭州, 310027
2.
北京大学信息学院, 北京, 100871
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
1560-8999 |
基金
|
国家教育部新世纪优秀人才支持计划
|
文献收藏号
|
CSCD:5845075
|