Calibration and Compensation of Anchor Positions for UWB Indoor Localization
Mingyang Qi, Bing Xue, Wei Wang
Abstract
The calibration of anchor positions is the prerequisite for high-precision ultrawideband (UWB) localization. However the second-order terms of anchor positions error and the distance constraint between anchors are neglected in existent calibration solution, which causes degradation of UWB positioning accuracy. In this article, a novel calibration and compensation algorithm of anchor positions based on tightly coupled UWB and micro-electromechanical system (MEMS) inertial sensor is presented. First, the trilateral localization model of UWB is established to analyze the impact of anchor positions error on UWB localization performance; it has been demonstrated that the positions error of UWB anchor encompass second-order error terms. To eliminate this error, the second-order measurement model of UWB/MEMS is built through the tight integration of UWB and MEMS, and the nonlinear error of anchor positions is estimated by the derivative unscented Kalman filter (DUKF). Then, the distance between anchors is calculated by the two-way ranging (TWR) strategy, the distance-constrained vector is exerted in DUKF to further improve the estimation accuracy of the anchor positions. Finally, experimental results validate the effectiveness of our proposed method, achieving high-precision calibration of anchor positions with a root mean square error (RMSE) of approximately 6 cm. This advancement significantly improves the adaptability and reliability of UWB localization systems.