Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters

Aquifer recharge and recovery systems (ARRS), which can broadly be analysed as seepage depth filters, in natural or engineered aquifers are gaining attention worldwide. Engineering predictions of their complex physical clogging behavior, however, continue to be challenging which has hindered the pre...

Full description

Saved in:
Bibliographic Details
Main Authors: Chew, Alvin Wei Ze, Law, Adrian Wing-Keung
Other Authors: School of Civil and Environmental Engineering
Format: Article
Language:English
Published: 2020
Subjects:
Online Access:https://hdl.handle.net/10356/144554
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-144554
record_format dspace
spelling sg-ntu-dr.10356-1445542021-02-09T08:01:22Z Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters Chew, Alvin Wei Ze Law, Adrian Wing-Keung School of Civil and Environmental Engineering Environmental Process Modelling Centre Nanyang Environment and Water Research Institute Engineering::Civil engineering Homogenization Theory Multi-scale Perturbation Analysis Aquifer recharge and recovery systems (ARRS), which can broadly be analysed as seepage depth filters, in natural or engineered aquifers are gaining attention worldwide. Engineering predictions of their complex physical clogging behavior, however, continue to be challenging which has hindered the predictive maintenance of these systems for energy and materials savings. To address this problem statement, we leverage the homogenization theory with the multiscale perturbation analysis as the feature engineering step to reduce the complexity of the physical clogging behavior in ARRS. The analytical approach systematically derives a unique homogenized representation which quantifies the clogging condition at the macroscale. A series of physical parameters are identified from the derived homogenized representation to build a pre-processed input layer into our own multi-layered neural network (NN) architecture for predictive analysis. Measured data extracted from the literature is then used to train and verify the NN model. The trained model yields an average error deviation of 20% between the model's predictions and the respective measurements for an optimized set of hyperparameters tested. We then discuss quantitatively how the model can be adhered to predict the timing for a concerned ARRS to reach its breakthrough stage for a range of operational conditions. Finally, we also demonstrate how the homogenized representation can be useful to determine an arbitrary filter's critical reaction rate and diffusion coefficient responsible for its breakthrough stage. Accepted version 2020-11-12T03:21:06Z 2020-11-12T03:21:06Z 2019 Journal Article Chew, A. W. Z., & Law, A. W.-K. (2019). Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters. Journal of Computational Science, 32, 21–35. doi:10.1016/j.jocs.2019.02.003 1877-7503 https://hdl.handle.net/10356/144554 10.1016/j.jocs.2019.02.003 32 21 35 en Journal of Computational Science © 2019 Elsevier B.V. All rights reserved. This paper was published in Journal of Computational Science and is made available with permission of Elsevier B.V. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Civil engineering
Homogenization Theory
Multi-scale Perturbation Analysis
spellingShingle Engineering::Civil engineering
Homogenization Theory
Multi-scale Perturbation Analysis
Chew, Alvin Wei Ze
Law, Adrian Wing-Keung
Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
description Aquifer recharge and recovery systems (ARRS), which can broadly be analysed as seepage depth filters, in natural or engineered aquifers are gaining attention worldwide. Engineering predictions of their complex physical clogging behavior, however, continue to be challenging which has hindered the predictive maintenance of these systems for energy and materials savings. To address this problem statement, we leverage the homogenization theory with the multiscale perturbation analysis as the feature engineering step to reduce the complexity of the physical clogging behavior in ARRS. The analytical approach systematically derives a unique homogenized representation which quantifies the clogging condition at the macroscale. A series of physical parameters are identified from the derived homogenized representation to build a pre-processed input layer into our own multi-layered neural network (NN) architecture for predictive analysis. Measured data extracted from the literature is then used to train and verify the NN model. The trained model yields an average error deviation of 20% between the model's predictions and the respective measurements for an optimized set of hyperparameters tested. We then discuss quantitatively how the model can be adhered to predict the timing for a concerned ARRS to reach its breakthrough stage for a range of operational conditions. Finally, we also demonstrate how the homogenized representation can be useful to determine an arbitrary filter's critical reaction rate and diffusion coefficient responsible for its breakthrough stage.
author2 School of Civil and Environmental Engineering
author_facet School of Civil and Environmental Engineering
Chew, Alvin Wei Ze
Law, Adrian Wing-Keung
format Article
author Chew, Alvin Wei Ze
Law, Adrian Wing-Keung
author_sort Chew, Alvin Wei Ze
title Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
title_short Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
title_full Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
title_fullStr Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
title_full_unstemmed Feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
title_sort feature engineering using homogenization theory with multiscale perturbation analysis for supervised model-based learning of physical clogging condition in seepage filters
publishDate 2020
url https://hdl.handle.net/10356/144554
_version_ 1692012984185913344