Factor Graph Optimization based Sensor Fusion Positioning Algorithm for UAV in GPS-denied Environment

In GPS-denied environments, especially in the dark and confined environments, traditional positioning methods are prone to performance degradation. Ultra-wideband (UWB) technology, which utilizes extremely wide bandwidth to achieve high-precision positioning, has attracted extensive attention under such circumstances. However, it also faces challenges such as measurement noise and accuracy degradation under non-line-of-sight (NLOS) conditions. To address the aforementioned issues, in this article, a fusion positioning method based on Factor Graph Optimization (FGO) that integrates IMU and UWB through data fusion is proposed. This approach ultimately achieves high-precision positioning for UAVs in their operational environments. According to the simulation results, by adopting the factor graph optimization, the proposed sensor fusion algorithm can achieve the median positioning error, the 95th positioning error, the root mean square error (RMSE), and the average standard deviation for RMSE with 0.0356 m, 0.0733 m, 0.0433 m, and 0.0197 m in the focused environments.