Online path planning based on Rapidly-Exploring Random Trees

This paper proposes an online path-planning algorithm based on Rapidly-Exploring Random Trees (RRT) applied to the autonomous navigation of a mobile robot. The proposed planner includes two heuristics to improve the performance and generates a set of collision-free paths, from which the one with the most similarity to a reference path given by a supervisor human operator is chosen. This reference can be given a priori when setting the start and goal positions, and be defined as the straight path between them. Simulations and experiments are made to evaluate the performance of the proposed method.