The complex and mixed traffic environment makes it a challenge for the widespread use of automated and connected vehicles (ACVs). It is necessary for these systems to have a better understanding of the traffic environment including interaction and gaming between multiple vehicles. In this study, a manoeuvre prediction and planning framework is proposed on the basis of game theories for complex and mixed traffic scenarios via vehicle‐to‐everything communication. In this framework, the interaction and gaming between multiple vehicles are considered by employing the extensive form game theories. In the payoff function, the risk assessment model based on trajectory prediction under uncertainty is employed to assess collision risks. Driving efficiency and preference are also combined in the payoff function. Uncertainty elements, including estimation and prediction, are considered to predict and plan by using Nash equilibrium of the extensive form game theory in mixed and behavioural strategies. Finally, this framework is applied and proved in different lane‐change scenarios. The results show that this framework could predict other vehicles’ driving manoeuvres and plan manoeuvres for ego vehicles by considering interaction and gaming between multiple vehicles, which helps ACVs understand the environment better and make the cooperative manoeuvre planning in complex traffic scenarios.


    Zugriff

    Download


    Exportieren, teilen und zitieren







    Decentralised spacecraft manoeuvre planning

    Cheng, Haoyang / Page, John / Olsen, John et al. | Emerald Group Publishing | 2014