Optimizing Motion Planning for Autonomous Vehicles: Path Planning, Curvature Analysis, and Steering Control
Francisco Badea Romero, Iñaki Echepare Soriano, Raffaele Marotta, Luca De Matteis · 2024
The paper introduces a comprehensive Motion Planning system tailored for autonomous vehicles, focusing on the critical aspect of global path planning within urban scenarios. It discusses the extraction of information from OpenStreetMap (OSM) to create a representative node network, followed by the generation of the shortest path using Dijkstra's algorithm. The path is then processed using B-Spline interpolation techniques for motion planning. Additionally, the analysis includes curvature evaluation and steering angle estimation based on Ackermann's relation. Road validation experiments and limitations in using geometric path analysis are also discussed.