A Real-Time RRT-APF Approach for Efficient Multi-Robot Navigation in Complex Environments
Year: 2025
Type: conference-paper
Abstract: This paper proposes a real-time multi-robot navigation method that integrates the Rapidly-exploring Random Tree (RRT) algorithm with the improved Artificial Potential Field (APF) approach. Since traditional path planning methods often face problems such as generating non-smooth paths and inefficient obstacle avoidance in changing environments, the RRT algorithm is used for initial path planning to pass through obstacles. Aiming to obtain a smooth collision-free path, Catmull-Rom spline smoothing is then introduced, which smooths the initially obtained trajectory and ensures that the curvature of the trajectory remains continuous. By combining the improved APF method, networked robots can then achieve safe navigation and effective obstacle avoidance in dynamic environments. The effectiveness of the proposed RRT-APF method is verified by both simulations and hardware experiments using real micro unmanned aerial vehicles.
Authors: Kaihuai Zhang, Mohsen Zahmatkesh, Martin Stefanec, Farshad Arvin, Junyan Hu
Institutions: Durham University, University of Graz
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4409659641