This paper presents an acoustic leak detection system for distribution water mains using machine learning methods. The problem was formulated as a binary classifier to identify leak and no-leak cases using acoustic signals. A learning methodology was employed using several detection features extracted from acoustic signals, such as power spectral density and time-series data. The proposed multi-strategy ensemble learning (MEL) approach demonstrates a significant improvement in performance, resulting in a reduction of false positives reports by an order of magnitude.
Editor's Note: Scranton Gillette Communications and the SGC Water Group are not liable for the accuracy, efficacy and validity of the claims made in this piece. The views expressed in this content do not reflect the position of the editorial teams of Water & Wastes Digest, Water Quality Products and Storm Water Solutions.