DISAGGREGATE TRAVEL DEMAND MODELS FOR THE SAN FRANCISCO BAY AREA. SYSTEM STRUCTURE, COMPONENT MODELS, AND APPLICATION PROCEDURES

Significant advances have recently been made in developing and applying disaggregate behavioral travel demand models to many aspects of urban travel decisions. What has not previously been developed is a full set of urban models integrated into a complete forecasting system for use by a metropolitan planning organization. The purpose of this paper is to describe the first such system, which was developed for the Metropolitan Transportation Commission, the designated metropolitan planning organization for the San Francisco area. First, the background of the current modeling project is briefly set out, followed by a description of the structure of the model system. The model development process--estimation, prediction testing, and validation--is described, and two computerized model application procedures--a regional network analysis system compatible with available urban transportation planning packages and a generalized policy analysis system based on random sample forecasting--are presented. Conclusions concerning the advantages and disadvantages of the system of disaggregate models are presented. /Authors/