CALIBRATION OF A PAST YEAR TRAVEL DEMAND MODEL FOR MODEL EVALUATION
David F. Pearson, P L Ellis, Stephen P Farnsworth
Abstract
David F. Pearson, P L Ellis, Stephen P Farnsworth
Abstract
Planning decisions for transportation infrastructure are based largely on forecast travel demand. To make informed decisions, the forecast travel demand must accurately reflect the response in traveler behavior and patterns of travel to changes in the attributes of the transportation system and the people using the system. Travel demand models are one of the tools used to estimate and forecast travel demand. The objective of this project was to perform an evaluation of the ability of the current modeling process in Texas to forecast travel behavior and patterns within the context of a controlled experimental design. The project attempted to calibrate travel demand models using travel surveys conducted in the San Antonio-Bexar County area in 1969. These models would then be applied to forecast travel behavior and patterns for the years 1980 and 1990, with the results compared to observed conditions. Due to data limitations, researchers did not consider the model developed for 1970 acceptable in terms of meeting the calibration criteria that were established. This negated the application of the model to 1980 and 1990 since the results could not be interpreted with any confidence. A number of recommendations on the model information needs for retention in future model calibration/validation efforts are made. The recommendations provide the basis for examining travel demand models within controlled conditions as a routine part of the modeling process.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Planning decisions for transportation infrastructure are based largely on forecast travel demand. To make informed decisions, the forecast travel demand must accurately reflect the response in traveler behavior and patterns of travel to changes in the attributes of the transportation system and the people using the system. Travel demand models are one of the tools used to estimate and forecast travel demand. The objective of this project was to perform an evaluation of the ability of the current modeling process in Texas to forecast travel behavior and patterns within the context of a controlled experimental design. The project attempted to calibrate travel demand models using travel surveys conducted in the San Antonio-Bexar County area in 1969. These models would then be applied to forecast travel behavior and patterns for the years 1980 and 1990, with the results compared to observed conditions. Due to data limitations, researchers did not consider the model developed for 1970 acceptable in terms of meeting the calibration criteria that were established. This negated the application of the model to 1980 and 1990 since the results could not be interpreted with any confidence. A number of recommendations on the model information needs for retention in future model calibration/validation efforts are made. The recommendations provide the basis for examining travel demand models within controlled conditions as a routine part of the modeling process.
Key concepts: Travel behavior, Context (archaeology), Demand forecasting, Process (computing), Operations research, Calibration, Computer science, Travel time