Machine Learning–Assisted Model for Leak Detection in Water Distribution Networks Using Hydraulic Transient Flows

2022 ◽  
Vol 148 (2) ◽  
Author(s):  
Amir Houshang Ayati ◽  
Ali Haghighi ◽  
Hamid Reza Ghafouri
Keyword(s):  
Machine Learning ◽  
Water Distribution ◽  
Leak Detection ◽  
Distribution Networks ◽  
Water Distribution Networks ◽  
Hydraulic Transient ◽  
Transient Flows

scholarly journals A Novel Leak Detection Approach in Water Distribution Networks

2021 ◽  
Author(s):  
Radhia Fezai ◽  
Kais Bouzrara ◽  
Majdi Mansouri ◽  
Hazem Nounou ◽  
Mohamed Nounou ◽  
...  
Keyword(s):  
Water Distribution ◽  
Leak Detection ◽  
Distribution Networks ◽  
Water Distribution Networks ◽  
Detection Approach

scholarly journals Extended-Horizon Analysis of Pressure Sensitivities for Leak Detection in Water Distribution Networks

2012 ◽  
Vol 45 (20) ◽  
pp. 570-575 ◽  
Author(s):  
Myrna Violeta Casillas Ponce ◽  
Luis Eduardo Garza Castañón, ◽  
Vicenç Puig Cayuela
Keyword(s):  
Water Distribution ◽  
Leak Detection ◽  
Distribution Networks ◽  
Water Distribution Networks

scholarly journals Model-based leak detection and location in water distribution networks considering an extended-horizon analysis of pressure sensitivities

2013 ◽  
Vol 16 (3) ◽  
pp. 649-670 ◽  
Author(s):  
Myrna V. Casillas Ponce ◽  
Luis E. Garza Castañón ◽  
Vicenç Puig Cayuela
Keyword(s):  
Water Distribution ◽  
Real Data ◽  
Leak Detection ◽  
Distribution Networks ◽  
Optimization Method ◽  
Least Square ◽  
Water Distribution Networks ◽  
Sensitivity Matrix ◽  
Model Based ◽  
Leak Detection And Location

In this paper, we propose a new approach for model-based leak detection and location in water distribution networks (WDN), which considers an extended time-horizon analysis of pressure sensitivities. Five different ways of using the leak sensitivity matrix to isolate the leaks are described and compared. The first method is based on the binarization approach. The second, third and fourth methods are based on the comparison of the measured pressure vectors with the leak sensitivity matrix using different metrics: correlation, angle between vectors and Euclidean distance, respectively. The fifth method is based on the least square optimization method. The performance of these methods is compared when applied to two academic small networks (Hanoi and Quebra) widely used in the literature. Finally, the three methods with better performance are applied to a district metering area of the Barcelona WDN using real data.

scholarly journals Experiments based comparative evaluations of machine learning techniques for leak detection in water distribution systems

2021 ◽  
Author(s):  
Maryam Kammoun ◽  
Amina Kammoun ◽  
Mohamed Abid
Keyword(s):  
Machine Learning ◽  
Distribution Systems ◽  
Water Distribution ◽  
Water Distribution Systems ◽  
Leak Detection ◽  
Distribution Networks ◽  
Machine Learning Techniques ◽  
Support Vector ◽  
Flow Data ◽  
Learning Techniques

Abstract Leakage in water distribution systems is a significant long-standing problem due to the huge economic and ecological losses. Different leak detection studies have been examined in literature using different types of technologies and data. Currently, although machine learning techniques have achieved tremendous progress in outlier detection approaches, they are still limited in terms of water leak detection applications. This research aims to improve the leak detection performances by refining the choices of learning data and techniques. From this perspective, commonly used techniques for leak detection are assessed in this paper, and the characteristics of hydraulic data are investigated. Four intelligent algorithms are compared, namely k-nearest neighbors, support vector machines, logistic regression, and multi-layer perceptron. This study focuses on six experiments based on identifying outliers in various packages of pressure and flow data, yearly data, seasonal data, night data, and flow data difference to detect leakage in water distribution networks. Different scenarios of realistic water demand in two networks from the benchmark dataset LeakDB are used. Results demonstrate that the leak detection accuracy varies between 30% and 100% depending on the experiment and the choices of algorithms and data.

Sign in / Sign up

Export Citation Format

Share Document