A step towards the live identification of pipe obstructions with the use of passive acoustic emission and supervised machine learning

Research output: Contribution to journalArticlepeer-review

Standard

A step towards the live identification of pipe obstructions with the use of passive acoustic emission and supervised machine learning. / Hefft, Daniel Ingo ; Alberini, Federico.

In: Biosystems Engineering, Vol. 191, 03.2020, p. 48-59.

Research output: Contribution to journalArticlepeer-review

Harvard

APA

Vancouver

Author

Bibtex

@article{548ab69195434719a954cc51e30f5a66,
title = "A step towards the live identification of pipe obstructions with the use of passive acoustic emission and supervised machine learning",
abstract = "A single passive acoustic emission sensor was used to collect signals coming from an obstructed pipe in a water recirculation system. Four geometrically different obstructions were investigated. The flow field of water around each obstruction was visualised with the use of 2D particle image velocimetry (PIV) to identify the different flow features. In parallel, the acoustic emission signals were acquired by locating a piezoelectric sensor on the outer wall of the pipe at the tip of the obstruction. The acoustic emission signals were then pre-processed and the frequency domain was extracted for 100 recordings in each case. Signals were processed further by using principle component analysis and a matrix is created for supervised machine learning algorithms. This methodology was applied over a range of four flow rates, all in fully developed turbulent flow. Results showed that different obstructions generated different acoustic signals and flow fields, which reflected the different flow fields observed with PIV. The average velocity and amplitude of the acoustic signals increased in magnitude with increasing flow rate. The machine-learning algorithm with highest prediction values was quadratic support-vector machine with predictions in the area of 95% accuracy or above. This makes the combination of machine learning and a single passive acoustic sensor a viable option to predict pipe obstructions and the type of obstruction. This may lead to a useful application for urban water supply or sewage systems as well as agricultural practice for field irrigation or the detection of nozzle blockages.",
keywords = "Obstruction, Hydrology, Pipe Blockage, Machine Learning, Particle Image Velocimetry, Acoustics",
author = "Hefft, {Daniel Ingo} and Federico Alberini",
year = "2020",
month = mar,
doi = "10.1016/j.biosystemseng.2019.12.015",
language = "English",
volume = "191",
pages = "48--59",
journal = "Biosystems Engineering",
issn = "1537-5110",
publisher = "Elsevier",

}

RIS

TY - JOUR

T1 - A step towards the live identification of pipe obstructions with the use of passive acoustic emission and supervised machine learning

AU - Hefft, Daniel Ingo

AU - Alberini, Federico

PY - 2020/3

Y1 - 2020/3

N2 - A single passive acoustic emission sensor was used to collect signals coming from an obstructed pipe in a water recirculation system. Four geometrically different obstructions were investigated. The flow field of water around each obstruction was visualised with the use of 2D particle image velocimetry (PIV) to identify the different flow features. In parallel, the acoustic emission signals were acquired by locating a piezoelectric sensor on the outer wall of the pipe at the tip of the obstruction. The acoustic emission signals were then pre-processed and the frequency domain was extracted for 100 recordings in each case. Signals were processed further by using principle component analysis and a matrix is created for supervised machine learning algorithms. This methodology was applied over a range of four flow rates, all in fully developed turbulent flow. Results showed that different obstructions generated different acoustic signals and flow fields, which reflected the different flow fields observed with PIV. The average velocity and amplitude of the acoustic signals increased in magnitude with increasing flow rate. The machine-learning algorithm with highest prediction values was quadratic support-vector machine with predictions in the area of 95% accuracy or above. This makes the combination of machine learning and a single passive acoustic sensor a viable option to predict pipe obstructions and the type of obstruction. This may lead to a useful application for urban water supply or sewage systems as well as agricultural practice for field irrigation or the detection of nozzle blockages.

AB - A single passive acoustic emission sensor was used to collect signals coming from an obstructed pipe in a water recirculation system. Four geometrically different obstructions were investigated. The flow field of water around each obstruction was visualised with the use of 2D particle image velocimetry (PIV) to identify the different flow features. In parallel, the acoustic emission signals were acquired by locating a piezoelectric sensor on the outer wall of the pipe at the tip of the obstruction. The acoustic emission signals were then pre-processed and the frequency domain was extracted for 100 recordings in each case. Signals were processed further by using principle component analysis and a matrix is created for supervised machine learning algorithms. This methodology was applied over a range of four flow rates, all in fully developed turbulent flow. Results showed that different obstructions generated different acoustic signals and flow fields, which reflected the different flow fields observed with PIV. The average velocity and amplitude of the acoustic signals increased in magnitude with increasing flow rate. The machine-learning algorithm with highest prediction values was quadratic support-vector machine with predictions in the area of 95% accuracy or above. This makes the combination of machine learning and a single passive acoustic sensor a viable option to predict pipe obstructions and the type of obstruction. This may lead to a useful application for urban water supply or sewage systems as well as agricultural practice for field irrigation or the detection of nozzle blockages.

KW - Obstruction

KW - Hydrology

KW - Pipe Blockage

KW - Machine Learning

KW - Particle Image Velocimetry

KW - Acoustics

U2 - 10.1016/j.biosystemseng.2019.12.015

DO - 10.1016/j.biosystemseng.2019.12.015

M3 - Article

VL - 191

SP - 48

EP - 59

JO - Biosystems Engineering

JF - Biosystems Engineering

SN - 1537-5110

ER -