基于排列熵的CEEMDAN分解算法研究OA

Complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)method has good decomposi-tion ability for complex signals.However,there are still mode confusion and too many pseudo components in CEEMDAN decomposi-tion results.To solve the above problems,this paper proposes a Permutation entropy based complete ensemble empirical mode de-composition with adaptive noise method(PECEEMDAN).This method embeds permutation entropy threshold detection in CEEM-DAN decomposition of the original signal,separates intermittent and noise signals,and directly performs empirical mode decomposi-tion(EMD)on the remaining signals.The experimental results of simulation signals and measured bearing signals verify that PECEEMDAN has better advantages in suppressing mode confusion,decomposition accuracy and fault feature frequency extraction.