An enhanced random noise suppression method for fiber optic gyroscopes to improve gyrocompass performance

Abstract

This paper presents a method for reducing random noise and zero-bias instability in the signal of a fiber optic gyroscope, based on a hybrid algorithm that combines an autoregressive mathematical model with an adaptive Kalman filter using a Sage sliding window. The results of the study demonstrate that this approach shortens the gyrocompass initialization time while maintaining the accuracy of defining initial orientation parameters, particularly the initial azimuth angle. Consequently, it enhances the overall performance of the gyrocompass in strapdown inertial navigation systems, especially in scenarios requiring rapid startup.