Feature Extraction in Pressure Signals for Leak Detection in Water Networks

Gamboa-Medina, Maria Mercedes; Reis, Luisa Fernanda Ribeiro; Guido, Rodrigo Capobianco

doi:10.1016/j.proeng.2014.02.075

Cited by 14 publications

(9 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A binary classifier, referred to as C4.5, has been used by Gamboa-Medina et al [22] to evaluate and to generate a set of rules to discriminate water leakages. In a controlled experimental laboratory circuit, water pressure data have been collected at high sampling rate (4 samples/s) by 15 sensors, producing 1-min sequences of 240 samples.…”

Section: State Of the Artmentioning

confidence: 99%

“…To our knowledge, most of them [19,20,[22][23][24]26] cannot be applied to residential scenarios, where only aggregated flow measurement records are available, i.e., only one sensing point. Specifically, the input data in Nasir et al [20] have been generated by means of the EPANET software [21] to simulate four pressure sensors and two flow sensors.…”

Section: Literature Reviewmentioning

confidence: 99%

“…In this case, non-faulty instances have been recorded by two pressure sensors, whereas the faulty instances have been simulated accordingly to that setup. Furthermore, in Gamboa-Medina et al [22], the data have been collected by an experimental circuit of 15 pressure sensors arranged along the pipe. Finally, in Alkasseh et al [24] and Mounce et al [19], both flow and pressure data have been collected from DMAs.…”

Section: Literature Reviewmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%

“…In particular, the experimental methodology adopted in Alkasseh et al [24], Gamboa-Medina et al [22], Nasir et al [20], Sanz et al [23] and Wan et al [26] is incompatible with the novelty detection paradigm [27,28], in which only the normal behavior is known. In Gamboa-Medina et al [22], the binary classifier has been trained to discriminate two classes: leak and no-leak. In Nasir et al [20], ANN and SVM have been trained to discriminate the leakage location and size; whereas Alkasseh et al [24] do not discriminate the presence of a leakage, but a linear correlation between the MNF (Minimum Night Flow) loss and the number of connections, the total length of the pipe network and the pipes' weighted mean age has been explicated.…”

Section: Literature Reviewmentioning

confidence: 99%

See 4 more Smart Citations

A Statistical Framework for Automatic Leakage Detection in Smart Water and Gas Grids

et al. 2016

View full text Add to dashboard Cite

Abstract:In the last few years, due to the technological improvement of advanced metering infrastructures, water and natural gas grids can be regarded as smart-grids, similarly to power ones. However, considering the number of studies related to the application of computational intelligence to distribution grids, the gap between power grids and water/gas grids is notably wide. For this purpose, in this paper, a framework for leakage identification is presented. The framework is composed of three sections aimed at the extraction and the selection of features and at the detection of leakages. A variation of the Sequential Feature Selection (SFS) algorithm is used to select the best performing features within a set, including, also, innovative temporal ones. The leakage identification is based on novelty detection and exploits the characterization of a normality model. Three statistical approaches, The Gaussian Mixture Model (GMM), Hidden Markov Model (HMM) and One-Class Support Vector Machine (OC-SVM), are adopted, under a comparative perspective. Both residential and office building environments are investigated by means of two datasets. One is the Almanac of Minutely Power dataset (AMPds), and it provides water and gas data consumption at 1, 10 and 30 min of time resolution; the other is the Department of International Development (DFID) dataset, and it provides water and gas data consumption at 30 min of time resolution. The achieved performance, computed by means of the Area Under the Curve (AUC), reaches 90% in the office building case study, thus confirming the suitability of the proposed approach for applications in smart water and gas grids.

show abstract

Section: State Of the Artmentioning

confidence: 99%