Automatic detection of burst in water distribution systems by Lipschitz exponent and Wavelet correlation criterion

•New equation to estimate the Lipschitz exponent in burst detection is proposed.•Wavelet correlation criterion and Lipschitz exponent are used in a new algorithm.•Burst detection resolution is improved.•By means of decision theory the probability of false positives is reduced. Bursts on water distri...

Full description

Saved in:
Bibliographic Details
Published in:Measurement : journal of the International Measurement Confederation 2020-02, Vol.151, p.107195, Article 107195
Main Authors: Trutié-Carrero, Eduardo, Cabrera-Hernández, Yosniel, Hernández-González, Arturo, Ramírez-Beltrán, Jorge
Format: Article
Language:English
Subjects:
Algorithms
Burst detection
Bursts
Correlation analysis
Criteria
Decision theory
High density polyethylenes
Lipschitz exponent
Water distribution
Water engineering
Water supply
Wavelet analysis
Wavelet correlation criterion
Wavelet transform
Wavelet transforms
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•New equation to estimate the Lipschitz exponent in burst detection is proposed.•Wavelet correlation criterion and Lipschitz exponent are used in a new algorithm.•Burst detection resolution is improved.•By means of decision theory the probability of false positives is reduced. Bursts on water distribution systems are a major problem, since they cause great loss of water, interrupt the supply and damage the streets and buildings. It is a challenge for water distribution systems to detect bursts as soon as they happen, reducing the probability of false positives and the size of the burst that can be detected. This paper solves this problem presenting a new equation that uses the Wavelet correlation criterion to estimate Lipschitz exponent and the theory of decisions to calculate the threshold used for burst detection. The algorithm was validated experimentally using 120 signals acquired from a high-density polyethylene pipeline. This paper shows that the algorithm and proposed equation obtain better results in burst detection than those reported by the scientific community.
ISSN:0263-2241
1873-412X
DOI:10.1016/j.measurement.2019.107195