基于形态特征的终端区进场中心航迹识别方法

Center Trajectory Extraction in Terminal Area Based on One Way Distance and Density-peak Clustering

为了挖掘终端区进场航空器交通流的分布特征,量化分析空中交通的复杂性,提出了一种基于多特征航迹相似度和密度峰值聚类(density-peak clustering,DPC)的中心航迹提取方法。首先,采用单向距离(one way distance,OWD)计算航迹之间的形状和物理距离,并结合空管实际运行航迹数据特征,考虑航迹之间的位置属性和航向属性,定义多特征航迹相似度模型。其次,使用密度峰值聚类算法对航迹数据进行聚类分析,提取聚类结果中每一簇中具有最高密度的真实轨迹作为中心航迹。最后,对双流国际机场终端区历史航迹数据进行实验分析,使用轮廓系数指标和基于密度的指标进行评价,并与层次聚类算法进行对比。结果表明,轨迹被划分为8个不同形态的类簇,该方法可以直观有效的识别出轨迹的整体运动特征并精确提取出真实的中心航迹。

In order to extract the central trajectory of approaching aircraft in terminal area,mine the space-time characteristics of flight trajectories,measure the degree of trajectory confusion,and analyze the complexity of air traffic,a central trajectory extraction method based on multi-feature trajectory similarity and density peak clustering was proposed,which used one-way distance to calculate the shape and physical distance between trajectories,and defines a multi feature trajectory similarity model taking into account the location attribute and heading attribute between trajectories.Through the analysis of density peak clustering,the real trajectory with the highest density in each cluster of clustering results was extracted as the center trajectory.Finally,the historical trajectory data of the terminal area of ZUUU Airport was analyzed,and the trajectories were divided into 8 clusters,using the silhouette coefficient index and density-based index,and then the accuracy and effectiveness of the method were verified by comparing with the hierarchical clustering algorithm.