摘要
根据统计数据挖掘思想提出了基于旅客订座记录的"初始需求"分解查找法。在分解查找过程中,同时考虑了机票预售阶段中的需求受截尾机理以及旅客短视型选择、策略型选择、取消预订和no-shows行为,并据此在舱位机票开放和关闭预订状态下,分别设计了历史真实"初始需求"与旅客初次购买首选"初始需求"量、取消预订量和no-shows量的关系算式。最后,使用模拟算例说明了所提方法的可行性和准确性:一方面,所提分解查找法能有效避免现有方法对航空网络"初始需求"量的高估问题;另一方面,所提无约束估计关系算式能有效控制现有分解查找法对航空网络旅客真实"初始需求"量的估计不足。
According to the statistical data mining,a decomposition search method for primary demand based on passenger name records was proposed.In the process of decomposition and search,the demand censoring mechanism in the pre-booking stage,as well as the myopia choice,strategic choice,cancellation of booking and no-shows behaviors of passengers were also considered.Accordingly,under the condition of both open and closed booking status,the formulas of the relationship between historical real primary demand and the first-choice primary demand,cancelled booking quantity,as well as no-shows quantity were designed respectively.Finally,numerical examples were given to illustrate the viability and the accuracy of the proposed methods.On one hand,the proposed decomposition search method can effectively avoid the overestimating problem of historical passengers’ primary demand which was caused by using both directly observation method and traditional decomposition search method.On the other hand,the underestimating problem of real network primary demand of historical passenger caused by the traditional decomposition search method could be controlled effectively by the proposed unconstraining formulas.
作者
郭鹏
周杰
GUO Peng;ZHOU Jie(School of Economics and Management, Guiyang University,Guiyang 550005, China;Business School, Sichuan Normal University,Chengdu 610101, China)
出处
《工业工程与管理》
CSSCI
北大核心
2019年第4期136-144,共9页
Industrial Engineering and Management
基金
国家社会科学基金一般项目(15BGL198)
国家自然科学基金青年基金资助项目(71601135)
教育部人文社科基金青年项目(16YJC630180)
贵州省教育厅高校人文社会科学研究自筹项目(2016ZC021)
贵阳市科学技术协会软科学研究项目(2016A01)
贵阳市创新型青年人才培养计划项目(2017经营管理及文化专门技术类)
2019年度贵阳市科学技术局-贵阳学院科技专项资金项目(GYU-KYZ(2019~2020)JG-25)
关键词
航空客运网络
需求无约束估计
旅客订座记录
分解查找
旅客策略行为
airline network
demand unconstraining estimation
passenger name records
decomposition search
strategic passenger behavior