Publication
Leak detection in water distribution networks based on water demand analysis
Conference Article
Conference
IFAC Symposium on Fault Detection Supervision and Safety for Technical Processes (SAFEPROCESS)
Edition
11th
Pages
679-684
Doc link
https://doi.org/10.1016/j.ifacol.2022.07.206
File
Abstract
This paper deals with the problem of leak detection in Water Distribution Networks (WDN). A leak detection method based on the water demand analysis of District Metered Areas (DMAs) is proposed. Historical leak-free data of water demand flow is used to extract minimum, maximum values, and statistical distributions of differences (errors) between demand flow and predicted values at different time hours of the day.
The concept of sensor fusion is applied to reduce measurement uncertainties. For this, a virtual measurement is generated that considers each hour of the day a feature and, combined, develops a more accurate error analysis capable of detecting leaks and estimating the leak size magnitude. Furthermore, to increase the accuracy of the leak detection method, prediction errors are analyzed in a moving time window. Finally, the performance of the proposed leak detection method is assessed by using actual data of different real DMAs of the Barcelona WDN.
Categories
artificial intelligence.
Author keywords
Water distribution network, leak detection, flow analysis
Scientific reference
D. Alves, J. Blesa, E. Duviella and L. Rajaoarisoa. Leak detection in water distribution networks based on water demand analysis, 11th IFAC Symposium on Fault Detection Supervision and Safety for Technical Processes, 2022, Pafos, Cyprus, Vol 55 of IFAC Papers Online, pp. 679-684.
Follow us!