A fusion positioning system with environmental-adaptive algorithm: IPSO-IAUKF fusion of UWB and IMU for NLOS noise mitigation

Abstract

Accurate positioning in non-line-of-sight (NLOS) scenarios persists as a critical challenge for ultra-wideband (UWB) systems. This paper proposes a collaborative positioning framework that integrates an inertial measurement unit (IMU). An improved particle swarm optimization and adaptive unscented Kalman filter (IPSO-IAUKF) algorithm based on environmental assessment is also designed. The threefold contributions include: (1) A tightly coupled positioning system architecture is constructed by deeply integrating UWB ranging with IMU motion measurements; (2) An improved particle swarm optimization (IPSO) algorithm is proposed to optimize the initial coordinate estimation of UWB using a dynamic inertia weight strategy; (3) An adaptive Unscented Kalman Filter (UKF) framework is designed, incorporating an environmental state discrimination threshold and a real-time noise matrix update mechanism to dynamically optimize the covariance matrix, thereby enhancing positioning robustness in complex noise environments. Multi-scenario trajectory simulations and practical experiments are conducted based on the established positioning model. Numerical simulation results demonstrate that the proposed fusion framework achieves a 52.6 % improvement in positioning accuracy compared to standalone UWB solutions, with a 44.6 % enhancement in noise resistance under NLOS interference compared to traditional fusion algorithms. Further practical tests reveal that the IPSO-IAUKF algorithm achieves average positioning accuracy improvements of 52.1 %, 45.5 %, and 46.0 % in two typical noise environments compared to conventional UKF and algorithms 1 and 2 used in this paper, respectively, while the maximum positioning error decreases by 44.6 %, 23.9 %, and 29.7 %, respectively. These results verify the superiority of this method in complex scenarios.