A dynamic cooperative lane-changing model for connected and autonomous vehicles with possible accelerations of a preceding vehicle

摘要

The emerging connected and autonomous vehicle (CAV) technologies offer a promising solution to design better lane-changing maneuvers that can reduce the negative impacts of vehicle lane-changing behavior on traffic operations. Existing studies on this topic have predominantly focused on designing lane-changing maneuvers for a subject vehicle (SV) and typically assumed that a vehicle in the target lane must decelerate to make space for the SV due to safety considerations. Nevertheless, jointly designing the trajectories of the SV and surrounding vehicles and allowing possible accelerations of a preceding vehicle may further alleviate the negative impacts of CAV?s lane-changing maneuvers. To investigate this possibility, this paper proposes a dynamic cooperative lanechanging model for CAVs with possible accelerations of a preceding vehicle. This model collects information of the surrounding vehicles and updates the lane-changing decisions for the SV in real time via three steps, namely lane-changing decision making, cooperative trajectory planning, and trajectory tracking. This model applies a linearized vehicle kinematic model to make lane-changing decisions for the SV given the states of the SV and surrounding vehicles, the minimum safety distance, and requirements on the comfort level for passengers. Furthermore, it dynamically designs the longitudinal and lateral trajectories for the SV and surrounding vehicles. Extensive numerical simulation experiments are conducted to evaluate the effectiveness of the proposed model. Results show that the proposed model increases the success rate of the SV?s lane-changing maneuvers, smoothens the trajectories of the SV and vehicles in the upstream direction at the cost of a slightly more significant oscillation of the last vehicle in the downstream direction. Overall, the proposed model reduces the negative impacts of lane-changing maneuvers on the surrounding traffic. The results also reveal the robustness of the model performance by varying several key input parameters in the experiments.