Title: An adaptive leak localisation system based on a multi-level analytics framework in piping network

Authors: Wen-Hao Png; Horng-Sheng Lin; Chang-Hong Pua; Mau-Luen Tham; Faidz Abdul Rahman

Addresses: Lee Kong Chian Faculty of Engineering and Science, Department of Electrical and Electronics Engineering, Universiti Tunku Abdul Rahman, Bandar Sungai Long, 43000, Kajang, Selangor, Malaysia ' Lee Kong Chian Faculty of Engineering and Science, Department of Electrical and Electronics Engineering, Universiti Tunku Abdul Rahman, Bandar Sungai Long, 43000, Kajang, Selangor, Malaysia ' Lee Kong Chian Faculty of Engineering and Science, Department of Electrical and Electronics Engineering, Universiti Tunku Abdul Rahman, Bandar Sungai Long, 43000, Kajang, Selangor, Malaysia ' Lee Kong Chian Faculty of Engineering and Science, Department of Electrical and Electronics Engineering, Universiti Tunku Abdul Rahman, Bandar Sungai Long, 43000, Kajang, Selangor, Malaysia ' Lee Kong Chian Faculty of Engineering and Science, Department of Electrical and Electronics Engineering, Universiti Tunku Abdul Rahman, Bandar Sungai Long, 43000, Kajang, Selangor, Malaysia

Abstract: Leak localisation is a growing concern in water distribution system (WDS). The conventional time-correlation analysis incorporates with acoustic sensing is a feasible leak localisation technique in single pipeline system. However, this technique is impractical and time-consuming in a piping network due to multi-directional transmission waves from the leak source. In this paper, we propose an adaptive leak localisation system incorporating a remote-acoustic sensor network and a multi-level analytics framework (MLAF) for piping networks. The MLAF overcomes the multi-directional waves issue in piping networks. The system is adopted with an automated flow control algorithm to ensure time-effective localisation without needs of human supervision. The performance of MLAF has been evaluated based on several emulated piping networks with various network topologies. The characterisation results demonstrated high adaptiveness and location accuracies. The excellent results of field prediction in a local district metered area (DMA) further validated the feasibility of the MLAF.

Keywords: pipeline monitoring; WSN; wireless sensor network; leak localisation; data analytics; path analysis; clustering analysis.

DOI: 10.1504/IJSNET.2021.117231

International Journal of Sensor Networks, 2021 Vol.36 No.3, pp.146 - 158

Received: 30 Jul 2020
Accepted: 03 Jan 2021

Published online: 13 Jul 2021 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article