A Continuous Curvature Rate Path Planner for Autonomous Cars in Narrow Environments
Domokos Kiss · 2025
Path planning is essential for autonomous vehicle navigation and has gained significant interest over the past decades. Many solutions address the curvature-bounded and continuous-curvature path generation for car-like vehicles, but only a few can effectively manage narrow environments, such as high-density parking and tight warehouse maneuvering, which often require complex maneuvers and reversals. This paper presents RTR-CCR, a sampling-based path planner that generates continuous curvature rate paths and excels in narrow planning scenarios. It employs two tree-structured graphs, similar to a bidirectional RRT planner, utilizing straight, circular, clothoid, and cubic spiral path primitives. We evaluate the performance of the algorithm in terms of path quality and running time, comparing it to other methods through Monte Carlo simulation experiments.