Theories and Confirmed Model of Urban Resident's Travel Demand: Considering Intra-household Interaction
柴彦威,张文佳
DOI: https://doi.org/10.3321/j.issn:0375-5444.2008.12.002
2008-01-01
Acta Geographica Sinica
Abstract:From the 1950s to the present, the research perspective of urban resident's travel demand transits from trip-based to activity-based, while the study units divide Traffic Analytical Zone (TAZ) into the individuals and even household. This paper summarizes that there are many problems of trip-based travel demand model in theory and practice. For this, the researches of activity-based and intra-household interaction come into being. Activity-based approach is the theoretical foundation of household-based travel behavior analysis, while its own foundation dated from the seminal work of Chapin (1974) and Hgerstrand (1970). Many literatures have confirmed that activity-based analysis could solve the problems using trip-based models so commendably that its contents extend and deepen from the 1980s to now. Thought most activity-based models still focus on the theories, household-based travel models gradually become the mainstream in transport planning in Western countries. Researches of household-based travel behaviors require to synthetically consider the effect on individual's activity-travel behavior of family attribute, the constraint to individual's behavior from household demand, and the couple constraint between household members. These aspects are briefly reviewed in this article. We discover that trip-based approach is hardly competent for studies of intra-household interaction. By contrast, activity-based approach could employ utility-based model, rule-based model, micro-simulation model and complex multi-variable model, such as structural equation model, to better interpret resident's activity-travel behaviors. Finally, Tianjin's time-use dairy survey data are used to confirm the theories of intra-household activity-based approach and to deeply understand the Tianjin's activity-travel behaviors. We model the socio-demographics as exogenous, and activity duration time and relevant travel time as endogenetic in the context of structural equation model (SEM). The final LIRSEL model fits well. The estimate results confirm that travel is derived of activity participation, and activity-based approach can better solve the problems of trip-based model, such as linking the discrete travels through activities interaction. Moreover, significance of activity-travel interaction between household heads positively exists. In particularly, males dominate the outside work-related activities and females dominate the outside household-related activities, and they jointly participate in non-work activities in Tianjin's households. Furthermore, comparing the total effect and direct effect on household heads' travel time of socio-demographics, which takes both the indirect effect of activities and other household members into consideration, we found the significance is not the same and positively argue that the total effect contains more actual information about activity-travel behavior. In conclusion, theories and confirmed model have proved that activity-based approaches are much better than the trip-based one, especially in terms of the intra-household interaction. In the future research, we should spread this perspective and approach to a better understanding of the behavior and transportation in transiting China.