Development of an advanced machine learning model to predict the pH of groundwater in permeable reactive barriers (PRBs) located in acidic terrain

Medawela, S; Armaghani, DJ; Indraratna, B; Kerry Rowe, R; Thamwattana, N

Development of an advanced machine learning model to predict the pH of groundwater in permeable reactive barriers (PRBs) located in acidic terrain

Medawela, S Armaghani, DJ Indraratna, B

Kerry Rowe, R Thamwattana, N

Permalink

Publisher:: ELSEVIER SCI LTD
Publication Type:: Journal Article
Citation:: Computers and Geotechnics, 2023, 161
Issue Date:: 2023-09-01

Closed Access

	Filename	Description	Size
	1-s2.0-S0266352X23003142-main.pdf	Published version	10.59 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Medawela, S
dc.contributor.author	Armaghani, DJ
dc.contributor.author	Indraratna, B https://orcid.org/0000-0002-9057-1514
dc.contributor.author	Kerry Rowe, R
dc.contributor.author	Thamwattana, N
dc.date.accessioned	2024-04-23T01:29:35Z
dc.date.available	2024-04-23T01:29:35Z
dc.date.issued	2023-09-01
dc.identifier.citation	Computers and Geotechnics, 2023, 161
dc.identifier.issn	0266-352X
dc.identifier.issn	1873-7633
dc.identifier.uri	http://hdl.handle.net/10453/178245
dc.description.abstract	This paper presents a comprehensive study of various Random Forest (RF) models for predicting increments of pH in groundwater treated by permeable reactive barriers (PRBs). This study uses data collected over 16 years from two PRBs installed in acid sulfate soils. The hyperparameters of the Random Forest have been optimised using three powerful optimisation algorithms: the Whale Optimisation Algorithm, the Tunicate Swarm Algorithm, and the Jellyfish Search Optimizer. The results show that the RF-Jellyfish Search Optimiser was the most accurate model for predicting pH increment in the downgradient. This model could predict pH increments for 80 data points as well as another PRB with a different reactive material. The model was also compared with a finite difference flow model from literature. The results showed that the RF-Jellyfish Search Optimizer was more accurate with the coefficient of determination and root mean square error values of 0.954 and 7.04, compared to 0.942 and 10.27 obtained from the finite difference model. Overall, the novel intelligent technique introduced in this study effectively predicts pH increments in treated groundwater and can also be applied to different PRBs.
dc.language	English
dc.publisher	ELSEVIER SCI LTD
dc.relation	http://purl.org/au-research/grants/arc/LP190100439
dc.relation.ispartof	Computers and Geotechnics
dc.relation.isbasedon	10.1016/j.compgeo.2023.105557
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	0905 Civil Engineering, 0914 Resources Engineering and Extractive Metallurgy, 0915 Interdisciplinary Engineering
dc.subject.classification	Geological & Geomatics Engineering
dc.subject.classification	4005 Civil engineering
dc.subject.classification	4019 Resources engineering and extractive metallurgy
dc.title	Development of an advanced machine learning model to predict the pH of groundwater in permeable reactive barriers (PRBs) located in acidic terrain
dc.type	Journal Article
utslib.citation.volume	161
utslib.for	0905 Civil Engineering
utslib.for	0914 Resources Engineering and Extractive Metallurgy
utslib.for	0915 Interdisciplinary Engineering
pubs.organisational-group	University of Technology Sydney
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology/School of Civil and Environmental Engineering
utslib.copyright.status	closed_access	*
dc.date.updated	2024-04-23T01:29:33Z
pubs.publication-status	Published
pubs.volume	161

Abstract:

This paper presents a comprehensive study of various Random Forest (RF) models for predicting increments of pH in groundwater treated by permeable reactive barriers (PRBs). This study uses data collected over 16 years from two PRBs installed in acid sulfate soils. The hyperparameters of the Random Forest have been optimised using three powerful optimisation algorithms: the Whale Optimisation Algorithm, the Tunicate Swarm Algorithm, and the Jellyfish Search Optimizer. The results show that the RF-Jellyfish Search Optimiser was the most accurate model for predicting pH increment in the downgradient. This model could predict pH increments for 80 data points as well as another PRB with a different reactive material. The model was also compared with a finite difference flow model from literature. The results showed that the RF-Jellyfish Search Optimizer was more accurate with the coefficient of determination and root mean square error values of 0.954 and 7.04, compared to 0.942 and 10.27 obtained from the finite difference model. Overall, the novel intelligent technique introduced in this study effectively predicts pH increments in treated groundwater and can also be applied to different PRBs.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/178245