Research on MEMS gyroscope motion signal processing method based on stochastic resonance and optimized mode decomposition

Abstract In order to process the motion signals of micro electro mechanical system (MEMS) gyroscopes more effectively, this paper proposes a method that combines tri-stable stochastic resonance (TSR) and optimal mode decomposition improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN). Firstly, we combined TSR with the crown porcupine optimization (CPO) algorithm and ICEEMDAN to improve the signal-to-noise ratio (SNR) of MEMS gyroscope motion signals. On this basis, the signals are decomposed into many intrinsic mode functions (IMFs). Secondly, the multi-scale permutation entropy (MPE) and dynamic time warping (DTW) are used to form the IMF component judgment criteria, which decompose these IMF components into noise, aliasing, and signal components. Then, Savitzky–Golay (SG) filter and wavelet packet threshold filter are used to filter the noise component and aliasing component separately, and the filtered results are superimposed with the original signal component to obtain the reconstructed signal. Finally, the proposed method is validated through simulation signals and measured motion signals from MEMS gyroscopes, and the results show its effectiveness and practicality.