Research on a small-noise reduction method based on EMD and its application in pipeline leakage detection
| Autor: | Laibin Zhang, Wei Liang, Xuchao Yu, Lu Wenqing |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
End effect
Computer science General Chemical Engineering Acoustics Noise reduction 020208 electrical & electronic engineering Energy Engineering and Power Technology 020206 networking & telecommunications 02 engineering and technology Management Science and Operations Research Computer Science::Numerical Analysis Industrial and Manufacturing Engineering Wavelet packet decomposition Maxima and minima Control and Systems Engineering Singular value decomposition 0202 electrical engineering electronic engineering information engineering Electronic engineering Safety Risk Reliability and Quality Food Science Leakage (electronics) |
| Zdroj: | Journal of Loss Prevention in the Process Industries. 41:282-293 |
| ISSN: | 0950-4230 |
| DOI: | 10.1016/j.jlp.2016.02.017 |
| Popis: | The noise included in pipeline pressure signal is a small noise whose energy takes a small proportion of pressure signal and is concentrated on high frequency components. However, it will influence pipeline leakage identification and even cause false alarms. Thus, a small-noise reduction method based on EMD (SNR-EMD) is proposed to remove small noise from pressure signal. EMD is applied for extracting the mean envelope of the signal. Then, small fluctuations around the mean envelope are considered to be small noises. Meanwhile, end effect of SNR-EMD is restrained by extrema mirror extension (EME). The results of simulation studies with SNR-EMD show that the larger the noisy signal's signal-to-noise ratio (SNR) is, the better noise reduction effect becomes. And SNR-EMD considered as a low-pass filter removes or reduces the high frequency components. Furthermore, superiorities of SNR-EMD are verified by comparison studies with wavelet packet transform (WPT) and singular value decomposition (SVD). Finally, a case study of leakage identification shows that SNR-EMD can improve the performance of leakage identification and reduce the possibility of false alarms, which makes much easier and further effective to distinguish the leakage mode from other modes after removing noise from pressure signal. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect