An Improved Hybrid A* for Efficient Path Planning in Simple Environments

Abstract

Path planning plays an essential role in many fields such as autonomous navigation, robot obstacle avoidance, and route optimization. Despite Hybrid A* being able to handle non-holonomic constraints and generating smoother paths than classical A*, it still suffers from computational efficiency issues, especially in relatively simple environments. By modifying parameters, this paper proposes an Improved Hybrid A* algorithm, which, while maintaining the kinematic feasibility of Hybrid A*, reduces the analytic expansion intervals, interpolation distances, and costs of switching direction. These parameters are tuned towards the reduction of reliance on heuristics and faster computation in relatively simple maze environments. The experimental results in maze environments with varying complexities demonstrate that, compared to the traditional Hybrid A* algorithm, the proposed Improved Hybrid A* algorithm can significantly enhance computing efficiency, particularly in less complex environments. This has shown that the morphology approach improves performance and is thus more suitable for autonomous navigation, where computational efficiency is important.