我国航班起降量中长期预测——基于GM(1,1)和LPGM预测的比较  

Forecast Medium-Long Term Air Traffic Flow Based on GM(1,1) and Logistics Population Growth Model

作　　者：李杨 LI Yang(Civil Aviation Telecommunication Development Co.,Ltd.,Beijing 100080,China)

机构地区：[1]民航电信开发有限责任公司

出　　处：《价值工程》2019年第16期185-189,共5页Value Engineering

摘　　要：本文对我国航班起降架次总量的中长期预测这一重要问题进行了研究。对于中长期预测模型,主要的挑战来自于数据量的缺乏,特别是我国航空数据积累的年度非常有限,大部分的预测和计量模型对样本量的要求,限制了这些模型的有效应用。另外,数据量的少也对中长期预测的效果评价带来了挑战,因此模型的解释能力往往是另外一个重点考虑的方面。本文我们分析对比了两个可以在小样本情形下进行长期预测的模型,灰色预测模型GM(1,1)以及Logistic群体增长模型(LPGM)。通过分析发现,LPGM模型的中长期预测要优于GM(1,1)模型,并且更为合理。根据LPGM模型预测,我国航空起降总量在2019-2030年平均年增长率约为6.9%,2021年将是发展的一个拐点,2030年航班起降量将达到年1900万架次,长期最大承载量极限约为2530万架次。This paper studies the important issue of the medium- and long-term prediction of the total number of takeoff and landing flights in China. For the medium and long-term prediction model, the main challenges come from the lack of data, especially the annual accumulation of aviation data in China are very limited. Most of the prediction and measurement models have requirements on the sample scale, which limits the effective application of these models. In addition, the lack of data also poses a challenge to the evaluation of the effect of medium and long-term predictions, so the explanatory power of the model is often another important consideration. In this paper, we analyze and compare two models that can be used to predict under a small number of samples, including long-term prediction, gray prediction model GM (1, 1) and Logistic population growth model (LPGM). The analysis found that the medium- and long-term prediction of the LPGM model is better than the GM (1,1) model and is more reasonable. According to the LPGM model, the average annual growth rate of China's aviation takeoff and landing in the period of 2019-2030 is about 6.9%, and 2021 will be an inflection point of development. The number of flights in 2030 will reach 19 million flights per year, and the limit is about 25.3 million.

关 键 词：航空流量 中长期预测 灰色模型 GM(1 1)模型 LOGISTIC增长模型 

分 类 号：V355.1[航空宇航科学与技术—人机与环境工程] TP183[自动化与计算机技术—控制理论与控制工程]