A Tightly Coupled Fusion Positioning Method Based on Particle Filtering With Effective Particle Neighbourhood Attraction Optimisation

Positioning and navigation systems have become indispensable infrastructure for modern society. However, standalone wireless positioning and inertial navigation systems each exhibit inherent limitations. Therefore, there is a strong need for integrated positioning techniques that can exploit the complementary strengths of multiple sensing modalities. Loosely coupled fusion approaches do not fully utilise raw observation information, whereas tightly coupled Kalman‐filter‐based methods remain limited in strongly nonlinear scenarios. To address these issues, this paper proposes a particle‐filter‐based fusion positioning method. In addition, an effective neighbourhood‐based particle set optimisation strategy is developed to mitigate particle degeneracy. Experimental results show that the proposed method significantly improves positioning accuracy over standalone positioning methods and tightly coupled Kalman‐filter‐based fusion methods. Compared with the cubature Kalman filter‐based fusion method, which provides the best performance among the conventional baselines, the proposed method improves positioning accuracy by 33.5%.