Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
of causality. A similar model is planned for the Bay Area system. JOINT ACTIVITIES LINKED EXPLICITLY ACROSS HOUSEHOLD MEMBERS Joint activities are those in which two or more household members travel together to and from an activity location and participate in the same activity while at that location. In the lower- level models, such as mode and destination choice, it is best to model such cases as a single joint deci- sion rather than as independent decisions made by differ- ent people. The Columbus and Atlanta model systems include models of household joint- activity generation and participation. The application of the Columbus model has shown that predicting joint travel can have significant implications for mode choice, so this type of model has 16 INNOVATIONS IN TRAVEL DEMAND MODELING, VOLUME 2 BRIEF DESCRIPTION OF CEMDAP (continued from page 15) TIME OF DAY, MODE, AND DESTINATION CHOICE MODELING SEQUENCE The work and school locations are predicted at the top level, while the work start and end times and work commute mode choice are modeled in sequence at the travel day level. The school start and end times are also predicted before the to- and from- school mode choice models. For all other activities, the tour mode is predicted at the tour level, followed by predicting the time of day and then the destination choice at the activity stop level. The departure time is derived from the predicted home stayâwork stay duration before each tour, activity duration atâ and travel time toâ each stop. NETWORK FEATURES, LEVEL OF SERVICE VARIABLES, AND MODELED TIME PERIODS CEMDAP can be used with any level of spatial reso- lution of zones and any number of time periods for level- of- service (LOS) variables. The DFW application uses a system of 4,784 traffic analysis zones for spatial representation and five time- of- day periods (a.m. off peak, a.m. peak, midday off peak, p.m. peak, and p.m. off peak) for LOS characteristics. No finer spatial units are used for land use variables. The effect of time- varying LOS characteristics is considered directly in work scheduling and indirectly in activity generation models through accessibility measures. The LOS attributes are also used in models of commute mode choice and nonwork- activity stop- location choice. Any time- of- day feature in CEMDAP is predicted in continuous time. The simultaneous prediction of work start and end times is currently implemented at 30-min periods, but it can be implemented at any finer time intervals. The school start and end times are predicted in continuous time by using hazard- based duration mod- els. The departure time for all other activities is also scheduled in continuous time. Available time windows are used in both the worker and nonworker scheduling models at the subpattern, tour, and stop levels. ACCESSIBILITY MEASURES Measures of accessibility from the home zone are used in activity generation models. The accessibility of a zone to another zone is calculated as the ratio of an attraction measure in the other zone relative to an impedance measure between the two zones (which is a function of travel times and costs). The parameters of these attraction and impedance functions were prede- termined from a destination choice model. The overall accessibility of a zone is then calculated as the average of the zone- to- zone accessibility measures. REFERENCES Bhat, C. R., J. Y. Guo, S. Srinivasan, and A. Sivakumar. Comprehensive Econometric Microsimulator for Daily Activity- Travel Patterns. In Transportation Research Record: Journal of the Transportation Research Board, No. 1894, Transportation Research Board of the National Academies, Washington, D.C., 2004, pp. 57â66. Guo, J. Y., S. Srinivasan, N. Eluru, A. Pinjari, R. Copper- man, and C. R. Bhat. Activity- Based Travel- Demand Analysis for Metropolitan Areas in Texas: CEMSELTS Model Estimations and Prediction Procedures, 4,874 Zone System CEMDAP Model Estimations and Proce- dures, and the SPG Software Details. Report 4080-7. Texas Department of Transportation, Oct. 2005. www.ce.utexas.edu/prof/ bhat/REPORTS/4080_7.pdf.