Q-Learning-Based Multi-Objective Global Path Planning for UGV Navigation

  • Hiago O. B. Batista UFV
  • Thayron M. Hudson UFV
  • Kevin B. de Carvalho UFV
  • Alexandre S. Brandão UFV

Abstract


This paper presents an offline path planning strategy for unmanned ground vehicles (UGVs) using Q-learning. The proposed method addresses path optimization in warehouse-like environments, where tasks involve item pickup and delivery to specific locations. The Q-learning algorithm trains an agent to determine the most efficient routes, with validation conducted in an 8×5 meter workspace equipped with an Optitrack motion capture system. The workspace was discretized into a 16×10 grid, allowing the Q-learning to effectively navigate through complex obstacle-laden scenarios. Experimental results indicate that the Q-learning approach outperforms traditional methods such as Dijkstra, A-star, and Breadth-First Search in terms of path length, number of turns, planning time, and overall success rate; being up to 7 times faster to plan a path and reducing the number of bends by up to 41%. The Q-learning based paths feature more linear segments, which contribute to energy savings and improved navigational efficiency. Future work will explore applications in heterogeneous multi-agent systems and enhancements in training time and agent collaboration.

Keywords: Training, Meters, Q-learning, Navigation, Path planning, Motion capture, Planning, Robots, Optimization, Multi-agent systems, Q-Learning, Artificial Intelligence, Autonomous Vehicles, Path Planning
Published
2024-11-09
BATISTA, Hiago O. B.; HUDSON, Thayron M.; CARVALHO, Kevin B. de; BRANDÃO, Alexandre S.. Q-Learning-Based Multi-Objective Global Path Planning for UGV Navigation. In: BRAZILIAN SYMPOSIUM ON ROBOTICS AND LATIN AMERICAN ROBOTICS SYMPOSIUM (SBR/LARS), 21. , 2024, Arequipa/Peru. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2024 . p. 48-53.