Modeling mixed traffic of automated and non-automated vehicles on highways at different speeds
Prof. Dr.-Ing. Perter Vortisch
Bundesanstalt für Straßenwesen (BASt)
PTV Transport Consult GmbH
PTV Planung Transport Verkehr AG
In recent years, transportation science has sought to assess the impact of connected and automated driving on traffic. Effects are expected at all levels, from travel demand to traffic flow. What all studies have in common is that they must be based on assumptions as to how the driving behavior of automated vehicles will actually be shaped. This is largely dependent on traffic and legal conditions, especially in mixed traffic when interacting with conventional vehicles. The introduction of automated vehicles brings a great opportunity to solve various problems in transportation. At the same time, however, studies show that mixed traffic in particular entails many risks.
The aim of the project is to develop recommendations of framework conditions for mixed traffic based on preliminary work and with the help of microscopic traffic flow simulation. These should be such that automated vehicles develop their potential on freeways as optimally as possible in terms of improving traffic safety, environmental impact and traffic flow.
Simulation models are built in PTV Vissim for selected highway sections and initially calibrated conventionally using macroscopic parameters. In order to improve the behavioral models in PTV Vissim, additional trajectory data will be analyzed. The necessary assumptions about the characteristics of automated vehicles in terms of driving behavior and sensor technology are developed with the support of experts from the field of automated driving development. Simulation calculations are then performed for defined scenarios under systematic variation of selected model parameters. The results of the simulation runs are examined with respect to traffic effects (including travel times and capacities), as well as with respect to effects on traffic safety and with respect to environmental effects. Since the driving behavior of at least the automated vehicles will change due to the possibility of communication with each other or with the infrastructure, we supplement the simulations with a model of this communication.