Innovations in Travel Demand Modeling, Volume 1: Session Summaries (2008)

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.

modeling route choice behavior by means of detailed GPS data obtained through PARROTS. Adopting knowledge from existing route choice models and cali- brating it for use in an activity- based context are also anticipated. Calibrating current models on real- world data will also be performed, along with improving com- putation time for large- scale simulations. The concepts currently used in Aurora, such as estimating S- curves as utility functions, estimating the effect of context vari- ables on maximum utility, evaluating the scheduling component, and extending learning facets, will be tested and evaluated. Additional concepts will be added, including the impact of life trajectory events, the impact of regular events, and strategic decisions. COMPREHENSIVE ECONOMETRIC MICROSIMULATOR FOR DAILY ACTIVITY- TRAVEL PATTERNS: RECENT DEVELOPMENTS AND SENSITIVITY TESTING RESULTS Chandra Bhat, Abdul Pinjari, Naveen Eluru, Ipek Sener, Rachel Copperman, Jessica Guo, and Sivaramakrishnan Srinivasan Chandra Bhat discussed Comprehensive Econometric Microsimulator for Daily Activity- Travel Patterns (CEMDAP), which is a continuous- time activity- travel prediction software currently being applied and evalu- ated in the Dallas–Fort Worth Metropolitan area. Vol- ume 2 provides a paper on the topic.2 The following points were covered in his presentation. • The development and testing of CEMDAP was funded by the Texas Department of Transportation (TxDOT). Janie Bynam and Bill Knowles of TxDOT and Ken Cervenka of North Central Texas Council of Gov- ernments provide assistance on the project. CEMDAP is based on a system of econometric models, with each model corresponding to the determination of one or more activity- travel attributes. The models are applied in a systematic sequence to generate the daily activity and travel patterns of all members of each household in the study area. • At a conceptual level, base- year inputs include aggregate sociodemographics, activity- travel environ- ment characteristics, policy actions, and model parame- ters. The synthetic population generator provides input to construct the detailed individual- level sociodemo- graphics for the base year. The socioeconomic, land use, and transportation system characteristics simulator (CEMSELTS) provides the sociodemographics and activ- ity environment. These characteristics link to the activity- travel simulator, CEMDAP, which generates individual travel patterns. These are loaded into a dynamic traffic assignment to develop link volumes and speeds. The link volumes and speed are fed back into CEMSELTS. • CEMDAP uses base- year inputs that include aggre- gate zonal- level demographic characteristics, land use patterns, the transportation network and level of service (LOS) measures, and any potential policy actions planned for a future year. The outputs for the forecast year include detailed activity- travel patterns. When the dynamic microassignment component is added, it will provide link volumes and speeds by time of day for the forecast year. • The modeling framework characterizes the activity- travel patterns of all household members, including adults, children, workers, nonworkers, stu- dents, and nonstudents. It explicitly considers space–time interactions and constraints. It models the allocation of maintenance activities, such as shopping, to household members and models parents’ escorting chil- dren to and from school. It generates and links joint activities of parents and children. CEMDAP adopts an interleaved approach to the generation of activity- travel patterns of all household members. It models 11 out- of- home activity purposes for adults and three out- of- home activity purposes for children. • The temporal resolution is a continuous time scale. The LOS data can be provided at any temporal resolu- tion. Five time- of- day periods are being used in the Dal- las–Fort Worth area application. The spatial resolution allows for any number of zones. The Dallas–Fort Worth application uses 4,874 zones. A standard Windows- based graphic user interface is used with CEMDAP. This interface allows users to modify model parameters and also provides a diagrammatic interface to help the user understand the logic of the system and the underlying models. • The CEMDAP software architecture allows for rapid implementation of system variants and expansions. Recent enhancements include the ability to model both adults and children incorporating spatiotemporal inter- dependencies, the incorporation of additional policy- sensitive variables to LOS, and the ability to process larger samples. • The synthetic population generator provides flexi- bility in how variables are aggregated. It supports differ- ent variable combinations to be synthesized and provides synthetic population for census tracks, block groups, or blocks. The synthetic population generator accounts for both household- and person- travel control totals. • CEMDAP was applied to examine a 10% and a 25% increase in in- vehicle travel times and a 10% and a 29ACTIVITY-BASED MODELS 2 See Bhat, C., A. Pinjari, N. Eluru, I. Sener, R. Copperman, J. Guo, and S. Srinivasan. Comprehensive Econometric Microsimulator for Daily Activity- Travel Patterns: Recent Developments and Sensitivity Testing Results. Volume 2, pp. 78–81.