Hybrid A∗ Based Motion Planning for Autonomous Vehicles in Unstructured Environment

Autonomous vehicles require a collision-free and comfortable motion trajectory at every time instant. It is a common method to generate a feasible path to the target state and append an optimization-based method for path postprocessing. In this paper a Hybrid A∗ based motion planning algorithm is presented for autonomous vehicles under unstructured circumstances. Firstly, the Hybrid A∗ algorithm is improved with a better heuristic function and a better search policy to realize a less-time consuming graph search in consideration of vehicle's motion model. Then, nonlinear optimization algorithm is applied to optimize the generated path further and realizes high quality of security and smoothness of the discrete trajectory. Finally, Catmull-Rom interpolation is combined to make waypoints continuous and easy-to-control. Simulation results concerning different tasks are described to demonstrate the validity of the proposed algorithm.