Development and comparison of choice models and tolling schemes for high-occupancy/toll (HOT) facilities

• We model the proportion of low occupancy vehicles using a high-occupancy–toll lane.
• We propose a new choice model based on value of time distribution in the population.
• We compare the proposed choice model to alternative models.
• Simple static and dynamic toll setting options are compared using the proposed model.

A high-occupancy/toll (HOT) lane is an increasingly popular form of traffic management strategy which reserves a set of freeway lanes for HOVs and transit users, while allowing low-occupancy vehicles (LOVs) to enter for a fee. In turn, HOT lanes maintain a minimal level of service by regulating the volume of entering LOVs. The focus of this paper is how to model the choice process of individual drivers, which dictates the volume of LOVs that choose to pay and take the HOT lane. Such models and the insights they provide can be very helpful for the toll setting process. Two simple formulations (an all-or-nothing assignment and an additive logit model) are compared with a proposed formulation based on the population value of time (VOT) distribution. Both static and dynamic toll setting algorithms are studied based on the proposed lane choice model, and their performance is compared under deterministic traffic behavior.