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.

and demographic characteristics, and changes in multi- modal transportation network characteristics. Other assessments might examine the impact of pricing policies, alternative work schedules, development patterns, transit fare changes, and other policy measures. The impact of new technologies on travel might also be examined. • Many of the current tour- based models use the tra- ditional zone- based spatial representation of a region and discrete time- of- day periods. Although this approach is expected to continue, activity- based models have the potential to provide a more continuous repre- sentation of the space–time domain. These models may also be better able to accommodate emerging behavioral paradigms and concepts. • Potential areas for further work and research include the development of guidelines for validation and assessing activity- based travel demand models, as well as guidelines for model comparisons. There is also a need for designing and conducting comprehensive experi- ments for performing controlled comparisons of activity- based travel model outputs and existing four- step model outputs under a variety of scenarios and policies. MODELING OF PEAK- HOUR SPREADING WITH A DISAGGREGATE TOUR- BASED MODEL Rebekah Anderson and Robert Donnelly Rebekah Anderson discussed the basic elements of the Mid- Ohio Regional Planning Commission (MORPC) model and the validation process. More detailed infor- mation on the development of the MORPC is provided in other sessions. Volume 2 provides a paper on the topic.3 The following points were covered in her presentation. • The MORPC model is a disaggregate tour- based model applied with the microsimulation of each individ- ual household, person, or tour. The model consists of nine separate models that are linked and applied sequen- tially. In order, these nine models are population synthe- sis, automobile ownership, daily activity pattern, joint tour generation, individual nonmandatory tour genera- tion, tour destination choice, time- of- day choice, tour mode choice, and stops and trip mode choice. • The choice mode hierarchy produces a record for every household and every person in the household. The time- of- day model is based on the “time windows” con- cept, accounting for the use of a person’s time budget over the day, with 16 hours available per person. It is a hybrid discrete choice departure time and duration model. The model has a temporal resolution of 1 hour for the modeled period between 5 a.m. and 11 p.m. The time- of- day model is applied sequentially among tours, with mandatory work, university, and school tours scheduled first. The model determines the departure time of each tour and the duration of the activity associated with the tour. The 190 departure and arrival time com- binations can be applied with relatively few variables. • In the development of the time- of- day model, a dis- aggregate validation was achieved using the Home Inter- view Survey (HIS) data records. The model results compared favorably with the observed values from the HIS. The time- of- day model has not yet been fully vali- dated against external data because MORPC does not have a sufficient number of traffic counts by peak hour or peak period. • The model area is divided into 1,805 internal and 72 external zones covering three counties and portions of four other counties. The validation process in 2000 focused on the entire model area and Licking County only. Examples of elements included in the validation were work trip distribution, volumes, and transit trips. The results from the MORPC model were compared to the 2000 Census Transportation Planning Package (CTPP). On a regionwide basis, the MORPC model esti- mated 660,031 work tours compared to 630,550 CTPP records— a difference of 4.7%. A comparison of district- to- district work tours revealed similarities and differ- ences between the MORPC model and the CTPP. The MORPC model estimate of all work trips to the central business district (CBD) was within 1% of the CTPP, whereas work trips in the North Corridor to the CBD were underrepresented by 5%. Regionally, the model overrepresented trips to Ohio State University (OSU) by 3%, but OSU trips in the North Corridor were overre- ported by 27%. • The MORPC model was also validated against traf- fic counts processed to represent directional average daily traffic for 2000. The criteria used to examine the accuracy of the MORPC model validation included the percent VMT error, the percent VMT root- mean- square error, and the percent volume root- mean- square error by facility type and volume group. The highway assignment validation was geographically structured by rings, sectors, and per districts. The results of this assessment indicate that model results are comparable to observed volumes. • Additional data and research are needed related to validating the time- of- day feature and other elements of the MORPC model for planning and policy analysis. Additional data on hourly traffic counts and vehicle clas- sification counts are needed for more extensive validation. Thomas Rossi moderated this session. 52 INNOVATIONS IN TRAVEL DEMAND MODELING, VOLUME 1 3 See Anderson, R. S., and R. M. Donnelly. Modeling of Peak- Hour Spreading with a Disaggregate Tour- Based Model. Volume 2, pp. 161–164.