Schuler period oscillation error suppression for inertial navigation systems based on reverse and wavelet transforms

In long-enduration pure inertial navigation of inertial navigation systems, the Schuler oscillation leads to a decrease in navigation accuracy. We propose a method based on reverse navigation and wavelet transform (RATF) to suppress Schuler oscillation errors. Reverse navigation solves the stored gyroscope and accelerometer data to predict the Schuler period in the forward navigation solution. Wavelet transform accurately identifies and extracts the instantaneous amplitude, frequency, and phase of the Schuler period predicted from the reverse navigation solution, thereby suppressing the navigation errors caused by the Schuler oscillations during the current forward navigation process. In the simulation and experimental data, we compare the RATF algorithm proposed in this paper with the pure inertial navigation algorithm and the internal damping network technology. The long-endurance positioning test results for vehicles and ships indicate that the RATF algorithm improves positioning accuracy by 50.77% and 67.28%, respectively, compared to the internal damping network technology. Therefore, the RATF algorithm not only better suppresses Schuler oscillation errors in the inertial navigation system, enhancing positioning accuracy during long-endurance navigation, but also does not rely on external sensor inputs, ensuring the autonomy of the inertial navigation system.