.png){kind=icon}
.png){kind=icon}
Return
Selective Noise Empirical Mode Decomposition
.png){kind=icon}
PRE
AI
DOI：10.1109/LSP.2025.3588082.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We propose selective noise empirical mode decomposition (SNEMD), an adaptive noise-assisted technique that enhances the performance of empirical mode decomposition (EMD) by introducing calibrated complementary noise and selectively extracting optimal modes. While existing noise-assisted methods mitigate mode mixing, they struggle with adaptive noise amplitude tuning and indiscriminate ensemble averaging. The latter often introduces mode mixing from individual realizations into the final result—a limitation overlooked in prior work. To overcome these issues, SNEMD introduces a novel mode mixing indicator (MMI) to discriminate the optimal modes from a collection of modes generated by EMD assisted with complementary noise of varying amplitudes. Theoretical analyses validate the effectiveness of the proposed MMI, while experimental results demonstrate that SNEMD delivers significantly improved decomposition performance, achieving at least a 14.47% gain in signal extraction accuracy compared to the best-performing existing method.
Keywords:
EEMD
noise amplitude
SNEMD
mode mixing
Journal
IF:
9.6
Papers:
1.1W
Citations:
1.7W