顾及地理环境的船舶行为识别研究  被引量:1

Ship Behavior Identification Considering Geographical Environment

在线阅读下载全文

作  者:周慧君 ZHOU Huijun(School of Geography Science,Nanjing Normal University,Nanjing 210023,China)

机构地区:[1]南京师范大学地理科学学院,江苏南京210023

出  处:《测绘地理信息》2023年第5期117-122,共6页Journal of Geomatics

基  金:国家重点研发计划(2018YFC1407405)。

摘  要:海洋地理环境时刻影响着航船的轨迹与行为,然而目前在海洋地理环境中利用轨迹对航船进行行为识别的方法却十分欠缺。针对上述问题,提出了一种顾及地理环境的船舶行为识别模型,该模型考虑到了航船轨迹本身的运动特性,且构建了船舶航行中所处的实时海洋环境,并将其转化为轨迹的地理环境语义,运用到船舶行为识别中。首先,根据船舶航行轨迹,利用海洋观测数据构建时空地理环境;然后,结合轨迹本身运动特征构建基于长短期记忆网络的船舶行为识别模型,并通过遗传算法思想对模型进行改进;最后,利用真实的船舶自动识别系统(automatic identification sys⁃tem,AIS)数据对模型的识别效果进行检验。结果表明:考虑航船所处的海洋地理环境可以有效提升对船舶行为识别的准确度,且所提模型比现有模型识别能力好。Marine geographical environment always affects the trajectory and behavior of ships.However,at present,there is a lack of methods to identify the behavior of ships using trajectories in marine geographical environment.To solve the problem,a ship behavior identification model considering geographical environment is proposed.The model considers the motion characteristics of the ship trajectories,and constructs the real-time marine environment in which the ship navigates,and then transforms it into the geographic environment semantics of the trajectories,which will be applied to the identification of ship behavior.Firstly,the spatiotemporal geographical environment is constructed by marine observation data according to the ship navigation trajectories.Then,a ship behavior identification model based on long short-term memory network is constructed with the motion characteristics of the trajectories,and improved by genetic algorithm.Finally,the identification effect of the model is tested by real automatic identification system(AIS)data.The results show that:considering the marine geographical environment where the ship navigates,the identification accuracy of ship behavior can be effectively improved;the proposed model has better identification ability than the existing models.

关 键 词:地理环境 地理语义 船舶行为识别 长短期记忆网络 海上智能交通 

分 类 号:P208[天文地球—地图制图学与地理信息工程] U675.79[天文地球—测绘科学与技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象