大面积航班延误实时预警研究

发布时间：2018-05-25 23:44

本文选题：大面积航班延误 + 实时预警　；参考：《中国民航大学》2017年硕士论文

【摘要】：随着我国航空运输业的迅速发展,空中交通容量不能满足日益增长的空中交通需求,加上天气、空管、机场等不确定因素,大面积航班延误频繁发生。大面积航班延误不仅造成巨大的经济损失,而且极易滋生不安全事件。实时、准确的对大面积航班延误进行预警有利于空管部门、航空公司和机场等相关单位及时、有效的调整运行方案,提高应急处置能力,确保航班运行安全、顺畅。本文首先基于大面积航班延误预警历史数据,利用多元统计分析方法探索大面积航班延误的时空分布规律,并用统计学方法验证主要延误因素的相关性;然后,提出基于航班计划的实时航班延误预警模型,评估区域管制空域实时延误程度;随后,采用传染病模型对大面积航班延误的传播机理进行研究,分析航班状态转化关系,建立了基于改进SIR的航班延误传播预警模型;最后,从大面积航班延误特征属性出发,再考虑实时的天气因素,应用支持向量机构建基于SVM的大面积航班延误等级预测分类策略,算例表明该模型的有效性,对比分析表明采用[0,1]归一化方式和径向基核函数时预警等级预测准确率最高,准确率高达94.8%。
[Abstract]:With the rapid development of China's air transportation industry, the air traffic capacity can not meet the increasing demand for air traffic. In addition to the weather, air traffic control, airport and other uncertain factors, a large area of flight delays occur frequently. Large-area flight delays not only cause huge economic losses, but also easily breed insecurity. Real-time and accurate early warning of large area flight delay is beneficial to air traffic control department, airline company and airport and other related units to adjust the operation plan in time and effectively, improve the emergency handling ability, and ensure the flight operation safety and smooth. Based on the historical data of large area flight delay warning, this paper uses multivariate statistical analysis method to explore the temporal and spatial distribution of large area flight delay, and uses statistical method to verify the correlation of main delay factors. A real-time flight delay warning model based on flight planning is proposed to evaluate the degree of real-time delay in regional control airspace. Then, the infectious disease model is used to study the propagation mechanism of flight delay in a large area, and the relationship between flight status transformation and flight status transformation is analyzed. A flight delay propagation early warning model based on improved SIR is established. Finally, the real-time weather factors are considered based on the characteristic attributes of large area flight delay. A large area flight delay grade prediction and classification strategy based on SVM is applied to support vector mechanism. An example is given to show the effectiveness of the model. The comparison and analysis show that the prediction accuracy of the early warning level is the highest when the [0Y 1] normalization method and radial basis function kernel function are adopted. The accuracy rate is as high as 94. 8%.
【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】：V355

【参考文献】