Path Planning Using Adaptive RRT*-Connect Algorithm Base on NURBS Curve Smoothing

Abstract

Optimal Path Planning for Robots refers to the autonomous search and planning of the shortest and smoothest path between the start and end points while avoiding obstacles in the space. This article presents an adaptive RRT*-Connect global path planning method based on NURBS curve smoothing (NARRT*-Connect). Building on the RRT*-Connect path planning, an adaptive step size adjustment strategy is introduced to accelerate the path search. In addition, NURBS curves are used for secondary path smoothing optimization, which does not require moving control points but only defines appropriate weighting criteria, ensuring that the path is smoothed while controlling local path changes within a safe threshold. Through simulation and experimentation, NARRT*-Connect is studied and compared with RRT*-Connect, RRT*, and other algorithms. The results indicate that the NARRT*-Connect algorithm can significantly improve iteration speed, generate shorter paths, and facilitate secondary safety smoothing of the paths.