Integrated framework of departure time choice, mode choice, and route assignment for large-scale networks

This paper presents a large-scale integrated modelling framework that can capture the relationships between travel mode choice, departure time choice and route choices simultaneously. Conventional transportation models have typically been applied to small scale networks to avoid the complexity of the large-scale simulation. While there are different approaches and strategies that researchers have presented to tackle the transportation congestion challenges, such strategies need to be accurately evaluated at the full system scale before implementation. The proposed framework integrates an econometric model for travel mode and departure time choice with a simulation-based dynamic traffic and transit assignment model. It addresses the interactions between the traffic and transit networks in addition to their interactions with the mode and departure time choice model. A case study of the proposed modelling framework is presented in this paper, where the effects of replacing the current flat transit fare used in the City of Toronto with a time-based transit fare structure are captured. It is found that an increase in the fare during the middle of the peak results in a reduction in transit vehicles crowdedness. The scenario analysis also shows that 1.85% of transit users will shift their departure time from the (congested) middle period of the peak to the (less costly) shoulders of the peak, whilst 2.6% of transit users optout of using transit in favour of either driving or accompanying another driver.