A Game-Theory and Risk-Field Based Automated Vehicle Motion Planning Method for Mixed Traffic Environments at Uncontrolled Intersections

With the development of technology, automated vehicles are beginning to appear in traditional multi-vehicle conflict scenarios such as uncontrolled intersections. The main challenge is to predict the behavior of other traffic participants. To address this issue and improve the traffic efficiency by optimizing the motion planning of automated vehicles (AVs) in MV-AV mixed traffic environments, a motion planning method based on game-theory and risk-field is proposed. Risk-field theory is used to build a driving behavior model of MVs at uncontrolled intersections, which AVs use to evaluate the driving behavior of other traffic participants. A game-theory framework assigns priority among AVs and MVs that conflict, considering their occupancy level of the conflict area and acceptable risk level. This method can obtain the optimal priority of vehicle in a linear calculation time while ensuring the driving safety in multi-vehicle conflict scenarios. The simulation results demonstrate the effectiveness of our motion planning method.