Multi-robot relative positioning and orientation system based on UWB range and graph optimization

Abstract

The relative position and orientation estimation is one of the most fundamental components for multi-robot systems, while it still remains an open and challenging research topic in infrastructure-free environments. In this paper, a novel graph optimization and UWB range based relative positioning and orientation estimation framework is proposed, which improves the estimation performance with a sliding window for multi-robot system in anchor-free and GNSS-denied environment. Single shot estimation capability of relative position and orientation benefits from the “single robot – multi tags” structure, which greatly improves the robustness and adaptability of the system. Real-site experiments demonstrates that our system can achieve centimeter-level relative positioning performance, and the influence of distance between nodes and the number of nodes on system performance is fully discussed through simulation, which lays a comprehensive preparation for future practical application.