Análise comparativa entre métodos de planejamento local para navegação de um robô móvel em ROS

Abstract

The goal of this work is to compare the performance of two obstacle avoidance navigation algorithms in the ROS (Robot Operating System) framework for the mobile robotics platform Turtlebot 3 Burger. The algorithms tested were DWA (Dynamic Window Approach) and TEB (Timed Elastic Bands). The analyzed criteria were geometric and angular precision and accuracy of the final pose, time and distance of the route, and usage of computational power. Both algorithms were tested on two different (symmetric and asymmetric) closed simulation environments with static obstacles in the Gazebo software, as well as on real environments with the same spatial arrangement. Both algorithms fulfilled the mission’s requirements, presenting the main difference in the usage of processing power.