Robust and Adaptive Calibration of UWB-Aided Vision Navigation System for UAVs

This letter proposes an adaptive calibration scheme that integrates Ultra-Wideband (UWB) anchor position estimation and adaptive path planning for Unmanned Aerial Vehicles (UAVs). A closed-form solution, independent of the initial guess of the UWB anchor positions, is rigorously derived, in the UWB anchor position estimation algorithm. Additionally, the covariance of the localization error is provided to adaptively evaluate accuracy. Based on the derived covariance, a path planning strategy is designed for the calibration phase. This strategy enables the UAV to adaptively plan its next best sampling point by optimizing the norm of the covariance matrix, with the convergence of the norm indicating the completion of the calibration phase. Simulation and experiments are carried out to verify the proposed method. Simulations with various parameters are implemented to compare the performance of different methods. In the experiment, a prototype system consisting of an easily deployable multi-UWB anchor base and a vision navigation enabled UAV is designed and realized. The system is then applied in the performance evaluation for the entire UWB-aided vision navigation task. The results of the simulations and experiments validate the effectiveness of the proposed adaptive calibration scheme. The integration of UWB anchor position estimation, path planning, and the vision navigation system leads to improved localization accuracy, efficient calibration completion, and enhanced performance in UWB-aided vision navigation tasks.