通航城市民航旅客吞吐量预测模型∗

摘要

民航旅客吞吐量的科学预测是确定机场建设规模和建设时机的重要前提条件。为构建适用于新建机场的客流预测模型，提高模型的预测精度，在考虑社会经济发展水平对通航城市民航旅客吞吐量影响的同时，引入航空网络结构的相关指标，比传统预测模型更为全面。对预测模型的不同构造形式进行了系统的比较分析，避免模型参数的多重共线性，保证模型设定无偏误。研究表明，通航城市的第三产业产值、旅游人数、平均路径长度和节点度均显著影响民航旅客吞吐量。所构建的预测模型对民航旅客吞吐量的解释能力达到了97．4％，对大部分城市的预测误差均在10％以内，具有较高的预测精度，但对某些旅游城市或边缘城市的预测精度较差。%Forecast of the passenger throughput of civil aviation is an important prerequisite for determining the timing and size of construction of airports.When proposing a forecasting model of passenger flow suitable for newly con-structed airports and improve the forecast accuracy of the traditional models,it should not only analyze the influences of social economy on passenger throughput of civil aviation,but also consider the factors,such as the structure of aviation network,etc.Meanwhile,in order to tackle the multicollinearity issue and ensure the accuracy of this model,a set of model forms are systematically compared.Results show that the passenger throughput of civil aviation in navigable cities is significantly affected by the output of tertiary industry,number of tourists,average path length and node degree.The explanatory power of the proposed forecasting model for the passenger throughput of civil aviation reaches 97.4%,and the forecast errors of most cities are within 10%.The proposed forecasting model has a high forecast accuracy.However, the forecast accuracy for some tourism cities or edge cities are found to be lower.