A Low-Cost UAV Swarm Relative Positioning Architecture Based on BDS/Barometer/UWB

The issue of relative positioning in unmanned aerial vehicle (UAV) swarms is critical for effective cooperative management and network construction. However, the current low-cost BeiDou system (BDS) exhibits poor accuracy, typically ranging from several meters to tens of meters, which fails to meet the basic requirements for cooperative positioning. To address this challenge, a combined positioning architecture based on BDS, barometer, and ultrawideband (UWB), alongside an accuracy optimization algorithm using gradient descent, is proposed. First, existing UAV swarm positioning technologies and algorithms are analyzed. By combining data from BDS, barometer, and UWB sensors, the accuracy problem of the combined system was transformed into an optimization problem. Then, models such as data credibility evaluation, information difference calculation, and gradient descent optimization were used to solve the optimization problem. Finally, a precision-optimized BDS/barometer/UWB-combined positioning architecture is proposed. Simulations and real-world experiments demonstrate that the fusion positioning architecture achieves higher cooperative positioning accuracy compared to mainstream systems, such as inertial navigation systems (INSs) and BDS. The final distance root mean square error (RMSE) is reduced to 1.313 m, a 69.0% improvement over the 4.231 m observed with standalone BDS.