Game theory and cooperative-competitive performances in real time traffic signal settings based on floating car data

The future of traffic control and management includes new technologies such as "connected" and "autonomous" vehicles. Connected vehicles will allow new traffic signal control systems that will be able to use information on vehicle positions (Floating Car Data, FCD). The use of FCD in traffic control systems can be beneficial by increasing sustainability of transportation in terms of fuel consumption reduction, reduced pollution and increased safety. However, not "connected" vehicles might not benefit or might even be damaged by the introduction of such systems. This paper intends to explore how an FCD-based adaptive traffic signal would impact different categories of vehicles in terms of cooperative-competitive performances. Game theory can be applied both in the implementation of algorithms and in assessing driver choices to participate in a cooperating and/or competitive signal regulation system, so a dedicated laboratory has been developed for testing the regulation algorithms and the interactions among drivers. The Nash bargaining equilibrium was used to implement a control algorithm which was tested with different percentages of "connected" vehicles. Some implementation challenges are discussed and new research directions outlined.