Simulation of Autonomous Path-Planning for Differential-Wheel Robots Based on A* Algorithm

Abstract

From designing to the employment of robots is time-consuming and costly, while running simulations before the manufacturing process could effectively eliminate unnecessary expenses of time and funding. This research discusses the integration of the A* path planning algorithm for autonomous navigation of differential wheeled mobile robots (DWMR) using Robot Operating System 2 (ROS2) with simultaneous localization and mapping (SLAM). The focus is to evaluate the system's performance through simulations in Gazebo and visualizations in RViz2. The method consists of simulating DWMR, combining the A* algorithm for pathfinding in maps generated using SLAM and LiDAR, and examining dynamic and kinematic modelling principles. The results show that in the comparison between the simulated path and the algorithm-generated path, the average deviation percentage is 2.697%. At this stage of research, autonomous path planning can only be completed within a well-discovered map. The significance of this study lies in demonstrating the feasibility of integrating optimized path planning with accurate mapping and localization for autonomous navigation. The findings validate the system’s efficiency and its potential applications in fields requiring precise autonomous navigation, such as manufacturing, and service robotics.