An Innovation Based DFA and EMD Method for Denoising Fiber Optic Gyroscope Drift Signal

One of tasks for how to reduce fiber optic gyroscope (FOG) random noise and drift is potentially useful for improving the precision of inertial navigation system (INS). In this paper, a novel signal denoising method is proposed that combines detrended fluctuation analysis (DFA) and empirical mode decomposition (EMD), named DFA-EMD. The noisy signal is first adaptively broken down into oscillatory components called intrinsic mode functions (IMFs) by EMD. Then DFA is used to measure each mode, which is independent of a comparison with the original signal. After that, a simple criterion based on DFA is designed to define the relevant modes for constructing the filtered signal. Experimental results, on simulated and real signals, show the superior performance of this proposed filtering over EMD-based denoisings and discrete wavelet threshold filtering.