APF-RRT*: Sampling-Based Motion Planning via Artificial Potential Field

Path planning is a fundamental problem in the field of autonomous systems, particularly when it comes to navigating complex environments and avoiding obstacles. Traditional sampling-based algorithms like Rapidly-Exploring Random Trees (RRT) and its variants such as RRT* have been widely used for solving high-dimensional path planning problems. However, these methods often suffer from suboptimal path quality and high computational costs in cluttered environments. To address these limitations, this paper proposes a novel path planning algorithm, APF-RRT*, which combines the advantages of Artificial Potential Fields (APF) with RRT* to enhance convergence speed, path smoothness, and overall trajectory quality. By integrating APF's attractive and repulsive forces into the sampling process, the proposed method guides the random search towards feasible regions while maintaining trajectory smoothness. Experimental results demonstrate that APF-RRT* not only reduces the number of required nodes but also improves the path quality compared to conventional RRT* and APF-RRT algorithms. APF-RRT* achieves a 15.1% reduction in average convergence time and a 52.2% decrease in node count while delivering smoother paths. The proposed algorithm shows promise in addressing path planning challenges in dynamic and complex environments, providing a reliable solution for real-time navigation tasks in autonomous systems.