数据驱动的终端空域进场运行评估方法研究

A Data-driven Method for Evaluating Arrival Operations in Terminal Airspace

终端空域交通稠密、结构复杂、冲突多发,因此终端进场运行在安全和效率管理中占据关键地位。充分融合指标、数据和方法,形成数据驱动的民航进场运行评估方法集。梳理相关国际组织及主流航空管理机构的民航运行效率指标框架,以此确立进场飞行时间和航迹复现性两个关键绩效指标。面向客观记录历史态势的雷达综合航迹,完成原始数据解析、关键信息提取,筛选出整体环境和系统条件相似的2个代表时段以供等效对比。由指标关注领域特征和对象数据性质,使用描述性统计与基于KDE的推论性统计评估进场飞行时间;结合特征选择,使用DBSCAN聚类评估航迹复现性。以长沙黄花机场为例验证研究了方法的有效性。

The terminal control area has dense traffic,complex structure,and frequent conflicts.The arrival operations of air traffic control(ATC)play a significant role in the management of safety and efficiency.Given this,this paper fully integrates the indicators,data,and algorithms to form a set of data-driven methods for evaluating the arrival operations of ATC in civil aviation.Firstly,the framework of ATC operational efficiency indicators of relevant international organizations and mainstream aviation management agencies was sorted out to establish two key performance indicators(KPIs):arrival flight time and the coherence of trajectories.Secondly,based on radar trajectories that objectively record historical operation situations,we completed the original data analysis and critical information extraction and screened out two representative periods with similar overall environment and system conditions for equivalent comparison.Thirdly,according to the characteristics of indicators′concerned areas and object data properties,descriptive statistics and KDE based inferential statistics were used to evaluate the arrival flight time;combined with feature selection,DBSCAN clustering was used to evaluate the coherence of trajectories.Finally,Changsha Huanghua airport was taken as an example to verify the effectiveness of the study.