Optimal Toll Setting Model Considering the Spread of Semi-Autonomous Vehicles with Variable Driving Modes Within a Trip

抄録

This study proposes an optimal toll-setting model considering the penetration of semi-autonomous vehicles (SAVs). These vehicles are manually driven on local roads and switch to autonomous driving on expressways. The model is formulated as a recursive logit model considering an optimization problem with equilibrium constraints for traffic assignment. The traffic assignment model is formulated as a fixed-point problem with the following features. Firstly, path selection criteria change based on driving mode. This is represented as a Markov chain based on link transitions. Human drivers are more short-sighted and make larger cognitive errors than SAV systems, expressed by the spatial discount rate and scale parameter. Secondly, expressway capacity varies with the mixing rate of SAVs. Thirdly, the value of time decreases under autonomous driving and fluctuates stochastically. Lastly, traffic demands are elastic and change with travel costs. A numerical calculation confirms the validity of the model and the solution algorithm.