基于PCMNL模型的多机场地区旅客机场选择分析  

Analysis of passenger airport selection in multi-airport regions based on PCMNL model

作　　者：宋溢露 羊钊[1] SONG Yi-lu;YANG Zhao(College of Civil Aviation,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)

机构地区：[1]南京航空航天大学民航学院,南京211106

出　　处：《哈尔滨商业大学学报（自然科学版）》2022年第1期107-112,共6页Journal of Harbin University of Commerce：Natural Sciences Edition

基　　金：国家自然科学基金(51608268)。

摘　　要：旅客对机场的选择影响着机场的发展,在多机场地区,地面交通方式多样,客运航线网络通达,旅客的选择行为更为复杂.针对多机场地区旅客机场选择决策行为及影响因素开展研究.利用预问卷确定选择机场的关键因素,依托某多机场地区典型实例,编制发放SP问卷,调查旅客在关键因素水平不同的情境下对机场的选择.基于回收的数据构建参数化随机选择集多项Logit(PCMNL)模型,最终模型命中率为88.3%,将模型计算结果与传统多项Logit模型相比,结果表明在挖掘到的显著因素相同时,前者精度优于后者.研究结果可用于测算各机场分担率,明确各机场定位.Passengers’ choice of airports affects the development of airports. In multi-airport regions, ground transportation methods are diverse and passenger route networks are accessible, and passengers’ choice behaviors are more complicated. This paper conducted research on the decision-making behavior and influencing factors of passenger airport selection in multi-airport regions. The pre-questionnaire were used to determine the key factors for choosing an airport. Relying on typical examples in a certain multi-airport region, SP questionnaires were distributed to investigate passengers’ choice of airports under different situations with different levels of specific key factors. Based on the recovered data, a parameterized probabilistic choice set multinomial Logit(PCMNL) model was constructed, and the final model hit rate was 88.3%. Comparing the calculation results of the model with the traditional multinomial Logit model, the results showed that when the significant factors mined are the same, the accuracy of the former is better than the latter. The research results can be used to calculate the share rate of each airport and clarify the positioning of each airport.

关 键 词：多机场地区 机场选择行为 PCMNL模型 SP调查 出行预测 LOGIT模型 

分 类 号：V351[航空宇航科学与技术—人机与环境工程]