LAUSR

Indoor positioning technology is crucial for enabling intelligent management and automation of indoor environments, such as smart buildings, factories and healthcare facilities. Ultra-wideband (UWB) navigation technology offers centimeter-level positioning accuracy. However, non-line-of-sight (NLOS) propagation significantly affects the positioning results of UWB. By combining UWB with inertial navigation system (INS) positioning estimation, better positioning performance can be achieved. This paper proposes a UWB and INS fusion positioning framework based on factor graph optimization (FGO). In this framework, when UWB is used for positioning, an efficient granular ball-based spectral clustering algorithm is first employed to cluster the measured distance values, reducing the impact of NLOS errors. This clustering algorithm improves the construction of the similarity matrix in traditional spectral clustering, reducing memory usage and minimizing errors caused by outlier measurements. An improved Taylor method based on the bacterial foraging optimization algorithm-Taylor is then used for the calculation of the moving target’s position. For the INS positioning system, inertial measurement unit pre-integration factors and bias factors are established. Finally, the INS and UWB information is fused using an FGO to reduce the effects of NLOS errors on UWB positioning outcomes and to minimize the cumulative errors of the INS. Simulation and experimental findings demonstrate that the proposed algorithm achieves improved performance.