摘要
文中聚焦出行者异质性问题,将潜在类别和混合Logit模型结合,构建了一种能同时实现用户细分和行为分析的选择行为模型,利用在线拼车选择意愿问卷数据进行模型标定.结果表明:出行者被细分为经济型、中立型和注重权益型三类,各类别间呈现出明显差异性;潜在类-混合Logit模型拟合度最优,相比对照模型模型,AIC值和BIC值分别降低3.5%和6.1%以上;在线拼车的特有属性补贴折扣对其选择概率影响最大,折数每降低一单位引起选择概率增加47%;选择其他出行方式的人群最易因出行费用的增加转向选择在线拼车,建议采取降低出行费用和加大拼车补贴力度的措施提高在线拼车选择概率.
Focusing on the heterogeneity of travelers,this paper combined potential categories with mixed Logit model to construct a choice behavior model that can realize user segmentation and behavior analysis at the same time.The model was calibrated by using the data of ridesplitting willingness questionnaire.The results show that travelers are subdivided into three categories:economical,neutral and rights-oriented,and there are obvious differences among them.The latent class-mixed Logit model has the best fitting degree.Compared with the control model,the AIC value and BIC value are reduced by more than 3.5%and 6.1%respectively.The special attribute subsidy discount of ridesplitting has the greatest influence on its choice probability,and every unit of discount will cause the choice probability to increase by 47%.People who choose other modes of travel are most likely to turn to ridesplitting because of the increase of travel expenses.It is suggested that measures should be taken to reduce travel expenses and increase subsidies to improve the probability of ridesplitting choice.
作者
方星雨
徐良杰
FANG Xingyu;XU Liangjie(School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan 430063,China)
出处
《武汉理工大学学报(交通科学与工程版)》
2025年第6期1224-1229,共6页
Journal of Wuhan University of Technology(Transportation Science & Engineering)
基金
湖北省重点研发计划项目(2023BAB022)。
