Vision-aided obstacle avoidance for the formation of multi-agent vehicles

Abstract

This thesis presents an in-depth study of vision-aided obstacle avoidance for multiagent formation control, integrating formation control, path planning, and the use of vision sensors. The paper thoroughly reviews the literature on coordination strategies, obstacle avoidance methods and sensor technologies in robotics. The main contributions of the thesis include the design and validation of formation controllers and obstacle avoidance algorithms, supported by simulations and experimental results with mobile robots equipped with an advanced vision sensor. Algorithms are validated in both MATLAB and Gazebo environments for their efficacy in the goal point navigation, narrow gap passing, and leader-follower obstacle avoidance. The research validates the robustness and adaptability of the proposed control algorithms through experiments, showcasing the ability to navigate towards designated goals while achieving precise formation configurations, encountering unforeseen challenges in real-world scenarios.