A new sampling strategy to improve the performance of mobile robot path planning algorithms

The path planning of a mobile robot using one of the sampling-based algorithms such as RRT(Rapidly-exploring random tree) and their variation RRT*, has received enormous attention especially in the last decade due to the ability of these algorithms to solve complex high-dimensional problems, as well as their completeness in probability i.e., they find a solution, if any, with infinite execution time. However, these algorithms have some limitations such as the low convergence speed for RRT* or producing a path far from optimal for RRT. This paper presents an improvement inspired by the A* algorithm, which can be applied to RRT and RRT* to overcome their shortcomings such as path cost and convergence time. Several simulation experiments were performed in different scenarios to compare improved RRT (A-RRT) and basic RRT, similarly for RRT*, these experiments showed a considerable reduction in run time and path cost.