Simulation Performance Comparison of A*, GLS, RRT and PRM Path Planning Algorithms

Abstract

Path planning is among the essential qualities of an autonomous robot. The ability to build a collision-free pathway from a pre-defined point to another is known as path planning. There are a variety of approaches offered, all of which vary depending on the search pattern and the map representation method. In this study, four robust path planning algorithms, namely: Probabilistic Roadmaps (PRMs), A-star, the Rapidly Exploring Random Trees (RRTs), and Generalized Laser Simulator (GLS), were simulated and their performance was measured and compared according to the total path distance covered, search time and path smoothness. The result obtained reveals that all the four algorithms could navigate and generate a feasible through the 2D map successfully. The GLS algorithm performs better in all the measured parameters followed by the PRM, RRT, and then the A* algorithm.