Résumé
Classical tools for measuring travel behavior, most of the time, are composed of household Origin-Destination surveys (revealed preferences) and related analysis methods aimed at developing knowledge about personal mobility (trip generation), mode usage (modal split techniques) or network performance (path choice and consequent trip assignment). It is the typical process of data collection for model development!
Based on a severely different perspective, the totally disaggregate approach --T.D.A.-- (Chapleau, 82-86) reversed the relationship between transportation travel data and models, introducing a process aimed at the integration of any territorial (zoning system, addresses, x-y coordinate locations), network (multimodal, nodes, links, lines) and demand (individual trips, counts, trip generators, personal characteristics, car ownership, household type, census data, trip purpose, trip chain, etc...) information within a unified framework. The key-elements of the T.D.A. were: 1) individual trip data processing throughout the analysis, where, incidentally, network modules such as access and path determination are adding information to the trip records, 2) usage of x-y coordinates -or monument and place declarations- as the basic spatially-referenced system for origin, destination, residential, intermodal junction locations.
With the T.D.A., urban transportation modeling benefits at numerous levels. Trip generation is addressed with socio-demographic variables; long-term travel forecasting is based on spatially-referenced demographic forecasts; modal shift modeling takes into account trip attribute changes and aggregate behavior; transportation network component usage is determined according to any kind of variable (residential zone, fare structure, modal costs, etc.). These applications have been already documented (Chapleau, 91-95).
Recent developments and applications of the T.D.A., such as activity map and trip generator analysis derived from origin and destination survey, are asking for a new conceptual framework. The best presently available language to describe the new methodology of the T.D.A. is the Object-Oriented Model.
The proposed paper will introduce and will demonstrate the concepts of OBJECTS, METHODS, PROPERTIES, CLASSES, INHERITANCE, ENCAPSULATION as applied in the context of urban transportation demand modeling. Origin and destination household surveys generate traditional objects of Households, Persons, Trips, but also produce other kinds of when specific logics are applied (trip chain validation, derived activity analysis). Furthermore, when integrating census databases, geographic information systems, transportation network geometry and schedules, land use data, a lot of new information are to be derived. Activity status, derived employment, trip generator behavior, spatial artificial desegregation are becoming new tools. Other perspectives like geographic sectors, transit or road network entities and attributes, will benefit of the new generalization of objects and methods.
Numerous applications, from the Montreal household origin and destination surveys of 1987 and 1993, conducted over a typical 5% sampling (a total of more than 800,000 individual trip records), are to be demonstrated, illustrating the role of the object-oriented concepts for developing new tools. In particular, data processing (objects, methods, properties) about personal mobility in geopolitical sectors, trip attraction for category generators, network analysis for transit riders and for vehicle interlining are typical cases where precise object-oriented methods and databases should be modeled. Moreover, the evolution of urban sprawl has been scrutinized and demonstrated when analyzing the behavior evolution of traveling workers in the Greater Montreal Area; every municipal sector, transport mode, age-gender cohort, bridge has been analyzed as objects.
In consequence, modeling urban transportation with an Object-Oriented methodology seems to be a promising way to integrate knowledge when that knowledge is disseminated with the actual information technology. |