Fiber optic gyroscope threshold denoising based on EMD-LWT

Abstract

Fiber optic gyroscope (FOG) drift data is often submerged in various noises backgrounds. It is very difficult to compensate for modeling drift signals directly. In order to better eliminate the noise mixed in the FOG temperature drift data, a hybrid EMD-LWT filtering algorithm based on empirical mode decomposition (EMD) and lifting wavelet transform (LWT) threshold denoising was proposed for gyro signals preprocessing. Firstly, the noise signal of fiber optic gyro is decomposed by EMD, and the noise term and the mixed modal term of the intrinsic mode functions (IMF) are judged according to the information entropy. Then the noise term is de-noised by LWT and the mixed modal term is denoised by wavelet transform (WT). A static test was performed on an interferential FOG to verify the effectiveness of the algorithm and compared with WT and LWT. The experimental results show that the proposed EMD-LWT filtering algorithm has better filtering effect. After processing, the root mean square error (RMSE) of the reconstructed signal is reduced by 63%, which effectively removes the noise in the FOG output.