Static LiDAR Assisted UWB Anchor Nodes Localization

摘要

The paper presents an indoor mapping and UWB (ultra-wideband) anchor localization method. The method, unlike most of the solutions described it the literature uses a static LiDAR (Light Detection and Ranging) mounted on a tripod rather than a robotic platform. It can be used at an any place, where employing a robot would be difficult (e.g. private homes), but since it requires manual LiDAR placement it would be most efficient in spaces of moderate areas. The proposed concept consists in mapping the environment of system installation while performing ranging measurements with deployed UWB anchor nodes. The SLAM (Simultaneous Location and Mapping) algorithm used for map integration and device localization relies only on LiDAR results. The matching is performed in two steps by finding an initial match based on corresponding landmarks extracted from the scans (intersections of the detected wall lines) and refining the results using an Iterative Closest Point algorithm. The anchors are localized based on the ranging results and SLAM-derived device locations using a Least-Squares based optimizer. The experiments have shown that the algorithm allows to construct a comprehensive map of the environment and localize the anchors with a root mean square error of 0.34 m, which is at similar level to analogous methods described in other works. The impact of anchor localization error on the systems performance was not significant. In both static and dynamic scenarios the difference in median errors obtained using reference and mapped anchors’ locations was about 0.05 m.