A Novel Local Trajectory Planning With Adaptive Extended Step Length for Autonomous Vehicles

This study proposes a novel and robust local trajectory planning for autonomous vehicles driving in open space. We improve the existing state space sampling local trajectory planning in order to utilize the higher-order information of the environment to generate safer trajectory and adaptively change the step length of each plan. At the same time, due to the incomplete consideration of the existing evaluation function we present a new evaluation function. The result of the simulation experiments shows that the new proposed local trajectory planning is more capable to plan safer trajectory than the existing method and is able to adjust the appropriate planning step length for each planning cycle according to the environment and the state of the autonomous vehicles.