TY - JOUR
T1 - Novel leakage detection by ensemble CNN-SVM and graph-based localization in water distribution systems
AU - Kang, Jiheon
AU - Park, Youn Jong
AU - Lee, Jaeho
AU - Wang, Soo Hyun
AU - Eom, Doo Seop
N1 - Funding Information:
Manuscript received April 20, 2017; revised August 4, 2017 and September 28, 2017; accepted October 5, 2017. Date of publication October 19, 2017; date of current version January 16, 2018. This work was supported in part by the Korea Ministry of Environment through “The Eco-Innovation project (Global Top project)” GT-SWS-11-02-007-9, in part by the Ministry of Education of the Republic of Korea, in part by the National Research Foundation of Korea under Grant NRF-2017R1D1A3B04034151, and in part by the Korea University Grant. (Corresponding author: Doo-Seop Eom.) J. Kang, S.-H. Wang, and D.-S. Eom are with the Department of Electrical and Electronics Engineering, Korea University, Seoul 02841, South Korea (e-mail: kanghead@korea.ac.kr; 08shwang@korea.ac.kr; eomds@korea.ac.kr).
Publisher Copyright:
© 2017 IEEE.
PY - 2018/5
Y1 - 2018/5
N2 - In many water distribution systems, a significant amount of water is lost because of leakage during transit from the water treatment plant to consumers. As a result, water leakage detection and localization have been a consistent focus of research. Typically, diagnosis or detection systems based on sensor signals incur significant computational and time costs, whereas the system performance depends on the features selected as input to the classifier. In this paper, to solve this problem, we propose a novel, fast, and accurate water leakage detection system with an adaptive design that fuses a one-dimensional convolutional neural network and a support vector machine.We also propose a graph-based localization algorithm to determine the leakage location. An actual water pipeline network is represented by a graph network and it is assumed that leakage events occur at virtual points on the graph. The leakage location at which costs are minimized is estimated by comparing the actual measured signals with the virtually generated signals. The performance was validated on a wireless sensor network based test bed, deployed on an actual WDS. Our proposed methods achieved 99.3% leakage detection accuracy and a localization error of less than 3 m.
AB - In many water distribution systems, a significant amount of water is lost because of leakage during transit from the water treatment plant to consumers. As a result, water leakage detection and localization have been a consistent focus of research. Typically, diagnosis or detection systems based on sensor signals incur significant computational and time costs, whereas the system performance depends on the features selected as input to the classifier. In this paper, to solve this problem, we propose a novel, fast, and accurate water leakage detection system with an adaptive design that fuses a one-dimensional convolutional neural network and a support vector machine.We also propose a graph-based localization algorithm to determine the leakage location. An actual water pipeline network is represented by a graph network and it is assumed that leakage events occur at virtual points on the graph. The leakage location at which costs are minimized is estimated by comparing the actual measured signals with the virtually generated signals. The performance was validated on a wireless sensor network based test bed, deployed on an actual WDS. Our proposed methods achieved 99.3% leakage detection accuracy and a localization error of less than 3 m.
KW - Ensemble convolutional neural network (CNN) and support vector machine (SVM)
KW - Leakage detection
KW - One-dimensional (1-D) CNNS
KW - Pipeline network localization
UR - http://www.scopus.com/inward/record.url?scp=85040767517&partnerID=8YFLogxK
U2 - 10.1109/TIE.2017.2764861
DO - 10.1109/TIE.2017.2764861
M3 - Article
AN - SCOPUS:85040767517
VL - 65
SP - 4279
EP - 4289
JO - IEEE Transactions on Industrial Electronics
JF - IEEE Transactions on Industrial Electronics
SN - 0278-0046
IS - 5
M1 - 2764861
ER -