AN AUTOMATIC CALIBRATION APPROACH FOR TOUR-BASED TRAVEL DEMAND MODELS: APPLICATION OF A DEMAND MODEL FOR THE GREATER CAIRO REGION

This study presents a novel approach to calibrating tour-based travel demand models, a crucial task in transportation planning. Unlike conventional trip-based models, tour-based models capture complex behavioral patterns, offering a more accurate representation of travel behavior. However, their calibration requires significant time and resources, particularly in developing countries with large populations. To address this challenge, we developed a genetic-based solution algorithm for automatic calibration of tour-based travel demand models. This approach integrates a bi-level mathematical model with transport modeling software, enabling the estimation of model parameters, such as mobility rate and destination choice parameters, that best fit the traveling behavior of residents. Our application of this approach to the Greater Cairo Region, using the most recent comprehensive Household Interview Survey, yielded promising results. The model validation metrics, including a coefficient of determination of 87%, a global relative error of 0.22, a Theil’s inequality coefficient of 0.26, and a point mean relative error of 36.2, demonstrate its effectiveness. This study contributes to the development of more accurate and efficient travel demand models, enabling the evaluation of transport policies and paving the way for similar applications in other large cities.