Author(s): Y. Yu; S. Rabab; R. Collins; T. Alps; E. Hampton; J. Boxall

Linked Author(s):

Keywords: Coustic sensing; Clustering; GMM; Leakage; Pipe material; Water pipe

Abstract: This study introduces a new data analysis methodology for the detection and location of leakage from water distribution infrastructure. A short-time Hanning windowing technique has been used as a low-pass filter for data segmentation and preprocessing. Our methodology capitalizes on the short-time zero-crossing rate and the short-time energy analysis of acoustic signals to effectively identify leakage. Additionally, Gaussian Mixture Modelling has been employed to cluster acoustic data, thereby improving the precision of leakage detection. We have conducted analysis and classification of leakage signals from both plastic and iron pipes, demonstrating advance over previous acoustic based detection method overcoming some issues wave attenuation in plastic pipes and enhance detection accuracy across different materials. Such advances are vital for both economic and environmental reasons.

DOI: https://doi.org/10.3850/iahr-hic2483430201-9

Year: 2024