摘要
对北京首都国际机场航班运行的历史数据进行了统计分析,建立了进离场小时架次及时间的三维数据结构,并利用DBSCAN算法进行聚类分析;在消除"噪点"的基础上,利用改进后的分位数回归算法生成分时段的具有95%概率的机场小时运行能力包线。仿真结果表明,各个时段的机场进离场架次具有一定的时间特性和差异:8点到10点处于离港早高峰,离港小时架次远大于进港小时架次;10点到22点进离港航班较为均衡,机场全负荷工作;其他时间段航班小时架次未达到最大容量值。
3D data structure of the arrival-departure hour sorties and time were established based on the statistical analysis of the historical data of capital airport flight operation,and proceeded the cluster analysis with DBSCAN algorithm. The airport hour operation capacity envelope with 95% probability was generated by using quantile regression algorithm on the basis of eliminating ＂ noise＂. Simulation results show that the arrival and departure sorties of the airport has certain amount of time characteristics and differences： at 8： 00-10： 00 am in the morning rush,departure hour sorties are far greater than arrival hour sorties; at 10： 00 am-22： 00 pm,arrival and departure flights is relatively balanced and full load working at the airport; other times flight hour sorties reduced and have not reached maximum capacity.
出处
《飞行力学》
CSCD
北大核心
2016年第4期86-89,共4页
Flight Dynamics
基金
国家自然科学基金资助(41501430
71571182
U1333116)
国家空管科研课题基金资助(GKG201405002)
中央高校基本科研经费资助(ZXH2012M003)
中国民航大学科研启动基金资助(2012QD02X)
大学生创新创业训练计划基金资助(201410059082)
教育部人文社科青年基金资助(14YJC630185)
关键词
机场容量
延误
分位数回归
聚类分析
airport capacity
flight delay
quantile regression
clustering analysis
