Research on leak location method of water supply pipeline based on negative pressure wave technology and VMD algorithm. (December 2021)

• Record Type: Journal Article
• Authors: Liu, Boxiang; Jiang, Zhu; Nie, Wei; Ran, Yijian; Lin, Hao
• Is Part Of: Measurement. Volume 186(2021)
• Publication Date: 2021-12
• Subjects: Pipeline leakage -- Negative pressure wave -- Variational mode decomposition -- Noise suppression -- Inflection point extraction -- Maximum peak -- Leakage location; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2021.110235
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: An adaptive denoising method for pipeline leakage signal based on VMD is presented. A novel NPW inflection point extraction method based on VMD is proposed. Noise in leak signal is effectively suppressed and the inflexion is exactly extracted. The proposed VMD-based method can accurately and stably locate the pipeline leakage. The presented method has tremendous application prospects in NPW-based technology. Abstract: Noise makes the inflection point of the negative pressure wave (NPW) difficult to be extracted. Therefore, an adaptive denoising method based on VMD (ADM-VMD) is proposed. ADM-VMD uses correlation coefficient to select effective intrinsic mode function (IMF) components, and obtains the optimal denoising signal according to the minimum information entropy. The traditional inflection point extraction method has the defect of poor robustness. Thus, an adaptive inflection point extraction method based on VMD (IPEM-VMD) is proposed. IPEM-VMD selects the optimal detail signal by kurtosis and extracts the inflection point according to the maximum peak. The experimental conditions with two pipe diameters were set up (0.08 m and 0.125 m). The results show that ADM-VMD can effectively smooth the noise and obtain a clear pressure drop curve. IPEM-VMD obtains a unique and clear maximum peak, which has good robustness to small noise. The average positioning error of the proposed method is 6.42%.