Research on Multi-Source Navigation Information Fusion Based on Adaptive Weighted Least Squares Algorithm Considering Signal Quality and Geometric Distribution

With the extensive application of satellite navigation technology in intelligent transportation and aerospace fields, the fusion of multi-source navigation information has become a key approach to solving the problem of high-precision positioning in complex environments. To address the issues of single weighting and insufficient robustness in traditional weighted least squares (WLS) algorithms for multi-source navigation information fusion in complex environments, this paper proposes an adaptive weighted least squares (AQG-WLS) algorithm based on signal quality and geometric distribution. This algorithm dynamically adjusts the weights of observations by constructing a composite weight model that integrates the carrier-to-noise ratio and the spherical geometric distribution of satellites. On the one hand, it uses the carrier-to-noise ratio to reflect the level of observation noise corresponding to signal strength. On the other hand, it introduces the area of spherical quadrilaterals to quantify the correlation of satellite geometric configuration, thereby avoiding the deterioration of the geometric dilution of precision (GDOP) caused by satellite clustering. Experimental results show that in scenarios with obstructions such as urban canyons, the AQG-WLS algorithm improves positioning accuracy by approximately 23% compared to the traditional WLS algorithm, and demonstrates stronger robustness against multi path effects and gross errors, providing an effective parameter estimation method for multi-source navigation fusion.