Research on PPP/INS Algorithm Based on Sequential Sage-Husa Adaptive Filtering

In the complex urban environment, the satellite navigation signal is polluted by gross errors due to external factors such as signal reflection, signal occlusion, signal electromagnetic. These factions cause the error of the satellite prior measurement noise random model based on satellite elevation angle in PPP/INS tight integrated navigation. In the traditional gross error detection method, the improper setting of the empirical threshold causes the traditional gross error detection algorithm to be insensitive to the small gross error of the measured value, which reduces the accuracy of the tight integrated navigation solution. In this paper, the sequential Sage-Husa adaptive filtering algorithm, combined with the upper and lower bounds of the measurement noise variance technology, effectively improves the non-positive definite problem of the measurement noise matrix, and guarantees the stability of the adaptive filtering algorithm. The algorithm estimates the filter adaptive factor in real time based on the measurement residual, and combining with the random model of satellite elevation angle, and the satellite elevation angle random model deviation caused by small pseudo range gross error is corrected in real time through the adaptive factor. Finally, the experiments show that when there are small gross errors in pseudo range observations, this algorithm is 2.7% higher than that of the traditional satellite elevation angle random model, the position accuracy is increased by 2.8%, 32.2%, 41.2%. Effectively improve the accuracy of the PPP/INS tight integrated navigation system in the complex urban environment.