摘要
出行方式和出行时间是决定出行需求时空分布的两个重要因素。为探讨出行方式对出行时间选择的影响,构建了出行方式和出行时间选择的离散-连续模型系统。其中,出行方式选择子模型应用了多项Logit模型,出行时间选择子模型则引入了持续时间模型。利用二元正态分布理论,通过关联两个子模型的误差项,建立了联合选择模型。以2013年某市居民出行调查中的通勤出行数据为基础,应用极大似然估计法标定了联合模型。研究结果表明,出行方式对出行时间有显著的影响,通勤者年龄、性别和收入等是影响出行方式和出行时间选择的重要因素。研究为预测出行选择行为和评价交通需求管理政策提供了新的思路。
Travel mode and trip timing are two important factors to determine the travel demand distribution in both temporal and spatial dimensions. To investigate the effect of travel modes on trip timings, a discrete-continuous model system of selecting different travel modes and trip timings is constructed. In this model system, the sub-models of travel mode and trip timing are predicted by a multinomial logit model and a duration model, respectively. According to bivariate normal distributions, a joint choice model is established by connecting error terms of these two sub- models. We then calibrate the joint model with maximum likelihood estimation method on a travel survey dataset containing the commuting trip records in an anonymous city of China in 2013. 2013 Our finding results demonstrate a significant effect of travel modes on trip timings, in which commuters＇ demographical information, including age, gender, and income, tends to be important factors for travel mode and trip timing selectior, These findings provide a new perspective to predict travel choice behaviors and also to evaluate travel demand management policies.
出处
《系统管理学报》
CSSCI
CSCD
北大核心
2017年第6期1131-1135,1142,共6页
Journal of Systems & Management
基金
国家自然科学基金资助项目(51478266)
上海市软科学研究计划重点项目(17692109300)
