An extended RRT approach for car-like systems

An extended method of the optimal rapidly exploration random tree (RRT*) for car-like robots is presented in this research paper. As it is known RRT* approach allows finding an asymptotic optimal solution to path planning problem. However, a lot of unnecessary waypoints are necessary as well as shar...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE Colombian Conference on Robotics and Automation (CCRA) pp. 1 - 6
Main Authors Navarro-Perez, Alvaro, Buitrago, Jaime
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2016
Subjects
Axles
Heuristic algorithms
Mobile robots
Robot kinematics
Trajectory
Online AccessGet full text

Cover

More Information
Summary:An extended method of the optimal rapidly exploration random tree (RRT*) for car-like robots is presented in this research paper. As it is known RRT* approach allows finding an asymptotic optimal solution to path planning problem. However, a lot of unnecessary waypoints are necessary as well as sharp corners between two consecutive segments. Also in this work, a removing redundant waypoints method is implemented by checking the visibility along a neighborhood from an initial configuration which will yield a pruned path. To avoid complex turning of the robot in each waypoint, a smoothness path method is implemented by using quintic Bezier spline as parametric curve taking into account the constraints imposed by the robot leading to increasing the convergence rate to reach an optimum or near an optimum solution and decrease the execution time. Finally, a simulation is presented to verify the effectiveness of the new path.
DOI:10.1109/CCRA.2016.7811416