A dynamic lane-changing trajectory planning scheme for autonomous vehicles on structured road

Autonomous driving has the potential of revolutionizing the mode of trip for human society with improved transport safety and accessibility and reduced emissions. Trajectory planning plays a fundamental role in autonomous driving implementation. This paper proposes a lane-changing trajectory planning method for autonomous vehicles on structured roads. The host vehicle collects information of the surrounding vehicles based on vehicle-to-vehicle (V2V) communication, and then a trajectory planning algorithm is presented to calculate a collision-free trajectory for lane changing. Accelerating, slowing down or returning to the original lane are possibly planned by the host vehicle in response to velocity changes of the surrounding vehicles. The quintic polynomial is used to represent the possible trajectories, and a cost function considering comfort and lane-changing efficiency is designed. The optimal trajectory is determined by solving a constrained optimization problem. The effectiveness of the proposed method is verified under different scenarios through simulations.