Self-Alignment for Inertial System in Vibration Environment Based on Improved EEMD

Abstract

Vibration is an important error source of inertial system. Eliminating vibration is important for improving the performance of initial alignment. In this paper, the improved Ensemble Empirical Mode Decomposition (IEEMD) and permutation entropy are used to eliminate the noises and the vibrations. Firstly, on the basis of the research on the influence of the frequency of artificial noise on the Empirical Mode Decomposition (EMD) decomposition, an improved EEMD is proposed to ensure that the vibration signal can be accurately decomposed. Second, the noise-related Intrinsic Mode Functions (IMFs), vibration-related IMFs and signal-related IMFs are separated according to the permutation entropy of each IMF. Finally, to achieve the purpose of eliminating noise and vibration interference, the signal-related IMFs are acted as the output of fiber optic gyroscopes and accelerometers. The experimental results show that the proposed method can effectively reduce the vibration and improve the efficiency and accuracy of the initial alignment.