基于航迹聚类的终端区进场程序管制适用性分析

ATC Serviceability Analysis of Terminal Arrival Procedures Using Trajectory Clustering

为了定量评估终端空域的空中交通管制服务水平,提出了描述进场飞行程序管制适用性的新概念。在对实际运行航迹数据特征分析的基础上,建立了基于对应雷达轨迹点逆向比对方法的航迹间相似性测度模型。应用层次聚类法对航迹数据集进行了聚类分析,提出了根据航迹聚类集构造平均航迹的方法。通过比较平均航迹和标准进场程序,建立了描述管制适用性的3个量化指标及其算法,包括:纵向偏离度、侧向偏离度和非常规航迹比例等。实例分析表明:以上指标能准确表示进场飞行程序提供空中交通管制服务的水平,并可以根据聚类结果来改进现有进场飞行程序结构。

In order to obtain quantitative evaluation of air traffic control（ATC） service level in terminal airspace, a novel concept is proposed to describe ATC serviceability of current terminal arrival procedures. Based on the characteristics analysis of real flight trajectories data, a flight trajectory similarity model is established by antidromic comparison between corresponding radar tracks. Trajectories are clustered through hierarchical clustering method. According to each trajectory cluster, a mean flight trajectory is constructed. Three metrics and its algorithms are set up through comparison between the mean trajectory and its standard arrival route, These ATC serviceability metrics include longitudinal de- viation,lateral deviation and outliers rate. Application results show that these metrics are able to indi- cate accurately the ATC service level of arrival procedures, and trajectories clustering results could be used to improve the current arrival procedures structure.