首都国际机场离港不正常航班特征分析与预测

发布时间：2018-06-05 12:09

本文选题：航班延误 + 不正常航班　；参考：《中国民航大学》2017年硕士论文

【摘要】：随着航空运输的发展,航班延误日益严重。本文从更微观的角度,研究离港不正常航班的延误特征,并根据其特征,建立离港不正常航班的延误预测模型,为不正常航班治理与应急预警提供可靠依据,以到达提高离港航班正常率的最终目的。首先,研究了离港不正常航班的延误特征。选取北京首都国际机场作为研究对象,在总结现有离港航班正常统计方法的基础上,提出了按照日、周、航空公司、离港方向和目的机场等新视角细化现有统计方法,重新审视离港不正常航班延误特征和影响因素与离港不正常率的相关性。另外,专门建立了归一化公平性指标,衡量离港航班正常性的差异。其次,基于各离港航班分布的不均衡性,研究了现有航班离港正常性统计指标,应用聚类方法,建立了基于离港方向的统计指标,并用时间序列预测模型验证了该指标的可行性。最后,根据离港不正常航班延误时序等特征,分别建立了时间序列预测模型-自回归滑动平均序列模型(ARIMA)和遗传算法优化的人工神经网络(BP)预测模型,预测离港航班延误。预测结果表明,遗传算法优化BP模型预测精度较高。本文提出了从离港方向的角度统计分析离港不正常航班,建立了衡量航班离港正常性的归一化指标和基于离港方向统计指标。这些指标和方法,不仅细化航班正常性统计方法,还提高了航班延误预测精度。
[Abstract]:With the development of air transportation, flight delay is becoming more and more serious. From a more microscopic perspective, this paper studies the delay characteristics of abnormal flights leaving Hong Kong and, according to its characteristics, establishes a delay prediction model for abnormal flights, which provides a reliable basis for the governance and emergency warning of abnormal flights. To achieve the ultimate goal of improving the normal rate of departure flights. First of all, the delay characteristics of abnormal departure flight are studied. This paper selects Beijing Capital International Airport as the research object, on the basis of summarizing the existing normal statistical methods of departing flights, proposes to refine the existing statistical methods according to the new perspectives of Japan, Zhou, airline, departure direction and destination airport, etc. Reexamine the characteristics of departure abnormal flight delay and the correlation between departure abnormal rate and influencing factors. In addition, the normalized fairness index is established to measure the difference of departure flight normality. Secondly, based on the uneven distribution of departure flights, the existing statistical indexes of departure normality are studied, and the statistical indexes based on departure direction are established by using clustering method. The feasibility of this index is verified by time series prediction model. Finally, according to the characteristics of departure abnormal flight delay time series, the time series prediction model-autoregressive moving average sequence model (ARIMA) and genetic algorithm optimized artificial neural network (BP) prediction model are established, respectively, to predict departure flight delay. The prediction results show that the prediction accuracy of BP model optimized by genetic algorithm is high. In this paper, a statistical analysis of departure abnormal flights from the departure direction is presented, and a normalized index and a statistical index based on departure direction are established to measure the normality of flight departure. These indexes and methods not only refine the statistical method of flight normality, but also improve the accuracy of flight delay prediction.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F562.8

【参考文献】