UC Irvine Institute of Transportation Studies

This paper summarizes the initial formulation of a micro-simulation model for activity-based travel demand forecasting that integrates household activities, land use distributions, regional demographics, and transportation networks in an explicitly time-dependent fashion. Intended to form the initial elements of an alternative to the conventional four-step transportation planning process, the prototype model incorporates an activity-based travel behavior model in a micro simulation approach utilizing a Geographic Information Systems platform to manipulate survey, demographic, land use, and network databases. 

An aggregate classification using travel diaries produces representative activity patterns which are specified implicitly in terms of temporal information, activity purpose, and sequencing. The classification also provides probability distributions of activity dimensions such as purpose and duration. Additional households are sampled and, based on demographic, land use, and network characteristics provided by the GIS, a target representative activity pattern is specified as are ambient activity densities. Activity characteristics such as purpose and duration are drawn from the distributions associated with the target pattern; trips are sequentially simulated based on a Monte Carlo approach of potential activity-specific destinations within a range of travel times from the prior and the home locations. The nature of the simulation is such that the simulated pattern, while maintaining the general characteristics of the target representative pattern, reflects the activity distributions and network characteristics of the household being simulated. The resultant set of activity patterns may be aggregated for any defined spatial-temporal limits. 

The model provides an activity-based method for estimating dynamic, linked-trip, origin destination demand matrices. Effectively replacing the generation and distribution components of the conventional process, the model represents a potentially important step toward the development of alternative transportation planning methods.