摘要
为缓解城市的交通拥堵,科学合理的引导小汽车出行者改乘公共交通出行显得尤为重要。以通勤出行者为研究对象,引入了有关价值观念的扩展心理潜变量,并建立了基于扩展TPB理论的潜变量MIMIC模型,分析了小汽车和公交出行方式下各潜变量之间的关系。然后,考虑随机森林算法有鲁棒性好、泛化能力强的优势,构建了带有心理潜变量的MIMIC-RF出行方式转移模型,通过实证结果得到有潜变量的模型比传统的模型具有更好的预测效果。因此,可借助该模型进行敏感度分析,结合关键影响因素的变化探究通勤者选择小汽车出行方式向公交转移的规律。
In order to alleviate the traffic congestion,it is particularly important to guide the car travellers to take public transportation.Taking commuter travellers as research objects,the extended psychological latent variable of the value concept is introduced,the latent variable MIMIC model is established based on the extended TPB theory,and analysis of the correlation is given for each latent variable between the car and public transport modes.Then,based on the advantages of robustness and generalization ability of the random forest algorithm,the MIMIC-RF travel mode transfer model with latent variables is constructed.The empirical results show that the model with latent variables has better prediction effect than the traditional model.Therefore,the model can be used for sensitivity analysis,and the changes of key factors can be used to explore the regular pattern of commuter travelers’ choice of public transportation instead of car travel mode during peak commuting.
作者
韩皓
谢天
HAN Hao;XIE Tian(School of Transportation,Shanghai Maritime University,Shanghai 201306,China)
出处
《内蒙古大学学报（自然科学版）》
CAS
北大核心
2020年第2期211-218,共8页
Journal of Inner Mongolia University：Natural Science Edition
基金
上海市人民政府决策咨询研究资助项目(2015ZD16B)。
