Visual-inertial navigation assisted by a single UWB anchor with an unknown position

作者：Haolong Luo, Danping Zou, Jiansheng Li, Ancheng Wang, Li Wang, Zidi Yang, Guangyun Li · 发表于：Satellite Navigation · 年份：2025 · DOI：10.1186/s43020-024-00153-6 · 被引用次数：12 · 研究领域：Robotics and Sensor-Based Localization、Indoor and Outdoor Localization Technologies、Inertial Sensor and Navigation

Abstract Navigation technology, which integrates vision, Inertial Measurement Unit (IMU), and Ultra-Wideband (UWB) sensors in GNSS-denied environments has gained a significant attention. However, inaccurate estimation of UWB anchor positions and improper sensor weighting among heterogeneous sensors significantly impairs the positioning accuracy and robustness of Visual-Inertial-UWB (VIU) systems. To accurately and rapidly estimate the UWB anchor positions, we employed the robust ridge nonlinear least-squares method to improve the accuracy and reliability of the estimated UWB anchor position. Additionally, we proposed a simple and effective method to assess the accuracy of the UWB anchor position using the geometric dilution precision principle, which facilitates rapid and accurate estimation of the UWB anchor position. Furthermore, we designed a method to calculate the estimated UWB anchor position error in real-world settings. Finally, we proposed a nonlinear optimization method with dynamically adaptive weighting based on the HELMERT variance component estimation principle, which assigns appropriate weights to heterogeneous sensors. To validate the feasibility and effectiveness of the proposed method, comprehensive simulations and real-world experiments were conducted. First, using Monte Carlo simulation and real-world experiments, we validated the effectiveness of the proposed methods for UWB anchor position and its accuracy estimation. Then, we conducted ablation experime...