A Study on Travel Demand Analysis accounting for Individual's Activity and Consumption Behavior

旅行行为与许多其他活动有很强的相互作用。在认识到活动参与和消费行为应该被考虑到精确的出行需求分析的基础上，建立了四个模型：1)两类自由裁量的时间分配双审查Tobit模型：该模型用于检验个人在工作日和非工作日对在家和在家外活动的时间分配。采用质量点法，引入一组误差分量来表示这种相关性。对模型的有效性进行了统计检验。2)基于收入约束下随机效用最大化假设的国内、国际旅行时间和支出预测的离散-连续选择模型系统由两个子模型组成。这些参数是使用女性的支出和旅行行为数据来估计的。结果表明，与非职业女性相比，职业女性的国内和国际旅游支出更高，个人年度国内和国际旅游支出更高。3)货币约束和时间约束下个人服务与商品消费的随机效用模型系统：该模型由非线性Tobit模型系统构成，其参数由个人的出行行为、时间使用和支出数据估计。参数估计表明，男性或高收入个体倾向于将更多的时间和金钱分配给非常规活动，例如观光或休闲活动。4)自由支配活动的货币和时间支出、地点和频率的同时选择模型系统：混合分布模型与Tobit模型相结合，估计未知参数。对系统中的模型组件进行了顺序估计。实证结果揭示了个人属性、旅游环境以及目的地属性对自由裁量活动的影响。少

Travel behavior has strong interactions with many other activities. Based on the recognition that activity participation and consumption behavior should be accounted for the precise travel demand analysis, four models were developed :1)A doubly-censored Tobit model of time allocation to two types of discretionary : The model is applied to examine individuals' allocation of time to in-home and out-of-home activities on working days and non-working days. A set of error components is introduced to represent this correlation, adopting a mass point approach. The validity of the model is examined statistically.2)A discrete-continuous choice model system which predicts the durations and expenditures of domestic and international travels : This model system is composed of two sub-models based on the assumption of random-utility maximization under income constraints. The parameters are estimated using women's expenditure and travel behavior data. The results indicate a strong relationship betwe … More en individual's annual expenditure for domestic and international travel, and higher expenditures for domestic and international travel of working women than those of non-working women.3)A random-utility model system of individual's consumption of service and goods under monetary and temporal constraints : The model is formulated as system of non-linear Tobit models whose parameters are estimated using individuals' travel behavior, time use and expenditure data. Parameter estimates indicate that male or high-income individuals tend to allocate more time and money to non-routine activities, for example sight seeing or leisure activities.4)A simultaneous choice model system of monetary and time expenditures, location, and frequencies of discretionary activities : Mixing distribution models were combined with Tobit models to estimate the unknown parameters. The model components in the system were estimated sequentially. Empirical results revealed the effects on the discretionary activities of the attributes of the individual, the travel circumstances as well as the attributes of the destination locations. Less