摘要
针对复杂运行场景中不易有效划分交通流的问题,提出了基于改进CURE(基于代表对象的聚类)算法的终端区航迹聚类方法。首先,采取等弧长插值重采样方法处理航迹的多维特征;然后,基于航迹多维特征改进相似性计算模型;最后,使用戴维森堡丁指数(DBI)自适应选取CURE算法的最佳聚类数目,并对终端区内航迹进行聚类。某机场终端区703条离场航迹的69763个数据点实例验证表明,该方法可在复杂运行场景中获取精确的交通流分布。
Aimed at the problem that it is difficult to classify traffic flow effectively in complex operation scenarios,an improved CURE(clustering using representative)algorithm is proposed to cluster the track in terminal area.Firstly,the multi-dimensional features of the track are processed by equal arc length interpolation resampling method.Then,the similarity calculation model is improved based on multi-dimensional features of the track.Finally,the optimal clustering number of CURE algorithm is selected adaptively with the Davies Bouldin index(DBI),and the tracks in the terminal area is clustered.An practical case of 69763 data points in 703 departure tracks in an airport terminal area shows that this method can obtain accurate traffic flow distribution in complex operation scenarios.
作者
纪新雨
初建宇
李印凤
傅子涛
李萌
JI Xinyu;CHU Jianyu;LI Yinfeng;FU Zitao;LI Meng(College of Civil and Architectural Engineering,North China University of Science and Technology,Tangshan 063210,Hebei,China;The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210023,China;China Civil Aviation Air Traffic Management Bureau in North China,Beijing 100621,China)
出处
《指挥信息系统与技术》
2021年第6期63-67,共5页
Command Information System and Technology
基金
民航华北空管局科技(201904,202002)
2020年民航安全能力建设(202072)资助项目。
关键词
层次聚类
插值重采样
相似性计算
hierarchical clustering
interpolation resampling
similarity calculation
