Abstract: In this paper, we propose a moving window-based Doppler smoothing pseudorange algorithm for multi-source fusion navigation systems where GNSS pseudorange observations are greatly affected by the environment, which results in a decrease in positioning accuracy. Firstly, we establish a GNSS/IMU/monocular vision tightly combined navigation system model based on the optimisation of the factor graph, and then we use open-source datasets to validate the algorithm and compare it with the conventional weighted Doppler smoothing pseudorange algorithm. and compared with the conventional weighted Doppler smoothing pseudo-ranging algorithm. The results show that the positioning accuracy of the algorithm in the three directions of E, N and U in pseudo-range positioning is 1.591, 2.892 and 2.001 m respectively, and the positioning accuracy is improved by 50.7 %, 61.7 % and 56.8 %. The positioning accuracy of the multi-source fusion navigation and positioning system is 1.390, 2.561 and 1.606 m, respectively, which is 9.8 %, 20.0 % and 11.1 % higher than the positioning accuracy in the three directions before smoothing. At the same time, compared with the conventional weighted Doppler smoothing, the positioning accuracy is improved by 4.7 %, 3.1 % and 0.5 % respectively.

Key words: multi-source fusion navigation, factor graph, pseudorange, moving window, Doppler smoothing