Research on Adaptive Kalman Filter Algorithm Based on GNSS/INS Integrated Navigation

Aiming at the problem that the GNSS/INS integrated navigation system is easily affected by climate and environmental noise in navigation and positioning, and it is difficult to determine the system state model, an adaptive Kalman filter algorithm based on innovation is proposed. The algorithm first uses the chi-square test to identify measurement outliers, and then adjusts the measurement noise with the innovation covariance to suppress outliers. The adaptive factor is used to adjust the system noise to improve the inaccuracy of the system noise covariance matrix. The AEKF algorithm proposed in this paper is verified through a sports car test. The test results show that compared with the EKF filtering algorithm, the AEKF algorithm proposed in this paper reduces the root mean square error of the east and north directions by 33.2 % and 51.5 %, and the root mean square error of the velocity by 35.0 % and 38.8 %, respectively. This proves that the algorithm can effectively suppress the state estimation error caused by measurement outliers, prevent filter divergence, and improve filter stability.