Abstract Robust and adaptive filtering algorithm has become one of the key technologies to realize integrated PNT because of its ability to control dynamic and observed anomalies in GNSS/INS integrated navigation. However, inappropriate decision model and robust factors will lead to further degradation of filtering. In order to solve this problem, a double-windows Sage-Husa robust algorithm and an improved decision model for the abnormal innovation ration (AIR) are proposed. The model anomaly is judged by the abnormal innovation ration decision based on loosely couple structure. We adopt the robust and adaptive filtering for the observed model anomaly and dynamic model anomaly respectively. This paper analyzes the measured data of urban avenues and streets in Wuhan. The experimental results show that the east, north and vertical position accuracy in complex environment is improved by 35.78%, 38.94%, 66.00%, and the root mean square error of position is 0.70 m, 0.69 m and 1.16 m.
GE Zhimin,JIANG Jinguang,ZHANG Chao et al. Application of Improved Robust and Adaptive EKF Algorithm in GNSS/INS Integrated Navigation[J]. jgg, 2023, 43(7): 740-744.
GE Zhimin,JIANG Jinguang,ZHANG Chao et al. Application of Improved Robust and Adaptive EKF Algorithm in GNSS/INS Integrated Navigation[J]. jgg, 2023, 43(7): 740-744.