Abstract: In order to distinguish the high frequency and low frequency modal components after VMD
( Variational Mode Decomposition) decomposition and improve the de-noising performance of VMD algorithm，a
de-noising method based on cloud similarity measurement is proposed． Firstly，the signal is decomposed by
VMD． The effective component and the noisy component are distinguished by calculating the cloud similarity
between each modal component and the signal． Then the noisy component is filtered by wavelet transform．
Finally，the denoised mode components and the effective component are reconstructed． Through simulation and
practical experiments，the proposed denoising method，VMD denoising method based on correlation coefficient
and VMD denoising method based on mutual information are used to process the noise signal． The SNＲ ( Signalto-
Noise Ｒatio) obtained by the proposed method is relatively high，which is 28． 214 1 dB． The mean square
error is relatively low，which is 6． 12 × 104，which verifies the superiority of the proposed method and the
feasibility of denoising the leakage signal of oil and gas pipelines．

Key words: cloud similarity, variational mode decomposition, signal de-noising