Vibration noise suppression method of SSA combined fading Kalman filter

The attitude estimation method based on singular spectrum analysis (SSA) combined with fading Kalman filter (FKF) is used to address the problem of low accuracy of attitude solution in vehicle navigation systems due to external vibration noise and maneuvering acceleration perturbation in microelectromechanical system (MEMS) accelerometer measurements. SSA preprocessing is first performed on the acceleration data to extract the trend and noise terms of the original signal, and then the FKF algorithm is used to fuse the SSA preprocessed acceleration signal with the gyroscope signal to output the estimated attitude. The results show that the SSA-FKF algorithm improves the accuracy of pitch angle and pitch angle estimation by 40% and 63.16%, respectively, when compared with the conventional Kalman filter algorithm (KF). The SSA-FKF solution can not only improve the attitude solution accuracy during external vibration, but also eliminate maneuvering acceleration interference.