基于集成模型的多维活动-出行决策研究

对个体活动-出行行为的5个决策维度进行联合建模分析，包括连续型的通勤距离、离散型的日通勤出行方式和通勤出发时间、小汽车拥有和日非工作活动量.一方面，作为内生变量，这5个决策维度受到个体性别、收入、教育水平等社会经济属性影响；另一方面，它们之间的相互影响关系也通过可见的结构关系及不可见的相关关系得到证实.比如，通勤出行距离越长，出行者家庭小汽车的拥有量越大；通勤出行距离、公交出行方式的选择均对日非工作活动量有负面影响；当通勤者家庭拥有更多的小汽车或者通勤出行距离很长时，他们选择小汽车和公交车的概率更高.本文证实了个体活动-出行决策维度之间的复杂作用关系，证实了对多维活动-出行决策进行综合集成分析的重要性.

Multi-dimensional Activity-travel Decisions Based on Integrated Model

This paper provides a unified modeling of five activity-travel decisions, including continuous commute distance, discrete commute mode and departure time, car ownership and number of non-work activities. It is found that on the one hand, these decisions are significantly influenced by individual traveler’s socioeconomic characteristics. The complex inter-relationships across behavioral decisions, on the other hand, are also confirmed. For example, (1) the longer the commuting distance, the more the private cars owned; (2) both commuting distance and choice of commuting by bus negatively influence the number of non-work activities; (3) individual has a higher propensity to commute by car and bus when more cars are owned by his/her family or the commuting distance is long. To sum up, this study confirms the complex interactions across individual’s activitytravel decisions, and emphasizes the importance of jointly modeling the multi-dimensional decisions.