Abstract:

In unstructured scenes， there are often obstacles of various sizes， and the path planning process that only considers obstacle avoidance methods such as detours will lead to decrease in vehicle traffic efficiency. For these problems， in this paper an intelligent vehicle path planning method with multiple obstacle-avoidance modes is proposed by integrating a layered collision detection strategy into the traditional Hybrid A* algorithm. Firstly， a double-layer grid map is constructed based on the vehicle chassis height， and a layered collision detection strategy is designed using the body contour and four-wheel contour. Then， through a well-designed heuristic function and cost function calculation method， the Hybrid A* algorithm can efficiently search for paths in multi obstacle scenes. Finally， the gradient descent method is used to smooth and optimize the path. Simulation and real vehicle experiment results demonstrate the effectiveness of the proposed algorithm in improving path search efficiency and significantly enhancing path smoothness. Moreover， the planned paths consider both crossing and bypassing strategies for obstacle avoidance， enabling vehicles to have better passability in multi-obstacle scenarios.

Key words: path planning, hybrid A* algorithm, hierarchical collision detection strategy, gradient descent method