Implementation of Classical Path Planning Algorithms for Mobile Robot Navigation: A Comprehensive Comparison

Abstract

Path planning for a mobile robot is the process of generating a suitable collision-free path to traverse between any two positions operating in challenging and cluttered environments. For efficient navigation and autonomous operations of mobile robots, intelligent and efficient algorithms of path planning play an important role. In this work, four classical path planning algorithms, namely, Dijkstra's, A-star <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathbf{A}^{\ast})$</tex> , rapidly exploring random tree (RRT) and artificial potential fields (APF) algorithms, have been explored. These algorithms are evaluated, simulated and implemented with a mobile robot with non-holonomic constraints. to test the merit of these algorithms, a suitable test environment has been set up with static obstacles. Based on the obtained map, the path planning algorithms are implemented and tested where the desired path was tracked by the mobile robot. to make a comparative study, standard performance indices have been evaluated for understanding the trade-off between these algorithms.