Systematic framework development for the construction of surrogate models for wastewater treatment plants

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

Surrogate modeling (also referred to as metamodeling) has attracted increased attention from researchers in various fields of engineering due to its use in computationally expensive engineering tasks such as Monte Carlo based global sensitivity analysis and process design optimization. However, the applications of surrogate models in the field of wastewater treatment modeling have not been extensively explored in the literature. In this work we present a systematic methodology for construction of powerful surrogate models to be used for the global sensitivity analysis of Benchmark Simulation Model 1 (BSM1) plant. A quasi-random design of experiments technique, Sobol sampling, is employed to generate an experimental design, which is further used to build surrogate models. A class of advanced metamodeling algorithms such as sparse polynomial chaos expansion (PCE) using least angle regression, Kriging interpolation, polynomial chaos Kriging (PCK), radial basis function (RBF) interpolation, multivariate adaptive regression splines (MARS), and a multilayer perceptron type feedforward neural network (ANN) are applied. The generalization error of the developed models has been estimated using holdout cross-validation with coefficient of determination (R2) and root mean squared error (RMSE) being used for evaluation of model predictive accuracy and model selection, respectively. The framework was further investigated forits suitability in a Monte Carlo based global sensitivity analysis using Sobol' method.The results obtained suggest that by following the framework, ANN and Kriging type surrogate models can effectively be constructed and used to estimate Sobol' sensitivity indices of WWTP design parameters.
Original languageEnglish
Title of host publicationProceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018
EditorsMario R. Eden, Marianthi G. Ierapetritou, Gavin P. Towler
Volume44
PublisherElsevier
Publication date2018
Pages1909-1914
ISBN (Electronic)978-0-444-64241-7
DOIs
Publication statusPublished - 2018
Event13th International Symposium on Process Systems Engineering (PSE 2018) - San DIego, United States
Duration: 1 Jul 20185 Jul 2018

Conference

Conference13th International Symposium on Process Systems Engineering (PSE 2018)
CountryUnited States
CitySan DIego
Period01/07/201805/07/2018
SeriesComputer Aided Chemical Engineering
ISSN1570-7946

Keywords

  • Surrogate modelling
  • Wastewater treatment modelling
  • Design of experiments
  • Metamodeling
  • Global sensitivity analysis

Cite this

Al, R., Behera, C. R., Zubov, A., & Sin, G. (2018). Systematic framework development for the construction of surrogate models for wastewater treatment plants. In M. R. Eden, M. G. Ierapetritou, & G. P. Towler (Eds.), Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018 (Vol. 44, pp. 1909-1914). Elsevier. Computer Aided Chemical Engineering https://doi.org/10.1016/B978-0-444-64241-7.50313-X
Al, Resul ; Behera, Chitta Ranjan ; Zubov, Alexandr ; Sin, Gürkan. / Systematic framework development for the construction of surrogate models for wastewater treatment plants. Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018. editor / Mario R. Eden ; Marianthi G. Ierapetritou ; Gavin P. Towler. Vol. 44 Elsevier, 2018. pp. 1909-1914 (Computer Aided Chemical Engineering).
@inproceedings{c4aadd3ec3934ef285c544cb2d088a66,
title = "Systematic framework development for the construction of surrogate models for wastewater treatment plants",
abstract = "Surrogate modeling (also referred to as metamodeling) has attracted increased attention from researchers in various fields of engineering due to its use in computationally expensive engineering tasks such as Monte Carlo based global sensitivity analysis and process design optimization. However, the applications of surrogate models in the field of wastewater treatment modeling have not been extensively explored in the literature. In this work we present a systematic methodology for construction of powerful surrogate models to be used for the global sensitivity analysis of Benchmark Simulation Model 1 (BSM1) plant. A quasi-random design of experiments technique, Sobol sampling, is employed to generate an experimental design, which is further used to build surrogate models. A class of advanced metamodeling algorithms such as sparse polynomial chaos expansion (PCE) using least angle regression, Kriging interpolation, polynomial chaos Kriging (PCK), radial basis function (RBF) interpolation, multivariate adaptive regression splines (MARS), and a multilayer perceptron type feedforward neural network (ANN) are applied. The generalization error of the developed models has been estimated using holdout cross-validation with coefficient of determination (R2) and root mean squared error (RMSE) being used for evaluation of model predictive accuracy and model selection, respectively. The framework was further investigated forits suitability in a Monte Carlo based global sensitivity analysis using Sobol' method.The results obtained suggest that by following the framework, ANN and Kriging type surrogate models can effectively be constructed and used to estimate Sobol' sensitivity indices of WWTP design parameters.",
keywords = "Surrogate modelling, Wastewater treatment modelling, Design of experiments, Metamodeling, Global sensitivity analysis",
author = "Resul Al and Behera, {Chitta Ranjan} and Alexandr Zubov and G{\"u}rkan Sin",
year = "2018",
doi = "10.1016/B978-0-444-64241-7.50313-X",
language = "English",
volume = "44",
pages = "1909--1914",
editor = "Eden, {Mario R.} and Ierapetritou, {Marianthi G. } and Towler, {Gavin P. }",
booktitle = "Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018",
publisher = "Elsevier",
address = "United Kingdom",

}

Al, R, Behera, CR, Zubov, A & Sin, G 2018, Systematic framework development for the construction of surrogate models for wastewater treatment plants. in MR Eden, MG Ierapetritou & GP Towler (eds), Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018. vol. 44, Elsevier, Computer Aided Chemical Engineering, pp. 1909-1914, 13th International Symposium on Process Systems Engineering (PSE 2018), San DIego, United States, 01/07/2018. https://doi.org/10.1016/B978-0-444-64241-7.50313-X

Systematic framework development for the construction of surrogate models for wastewater treatment plants. / Al, Resul; Behera, Chitta Ranjan; Zubov, Alexandr; Sin, Gürkan.

Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018. ed. / Mario R. Eden; Marianthi G. Ierapetritou; Gavin P. Towler. Vol. 44 Elsevier, 2018. p. 1909-1914 (Computer Aided Chemical Engineering).

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

TY - GEN

T1 - Systematic framework development for the construction of surrogate models for wastewater treatment plants

AU - Al, Resul

AU - Behera, Chitta Ranjan

AU - Zubov, Alexandr

AU - Sin, Gürkan

PY - 2018

Y1 - 2018

N2 - Surrogate modeling (also referred to as metamodeling) has attracted increased attention from researchers in various fields of engineering due to its use in computationally expensive engineering tasks such as Monte Carlo based global sensitivity analysis and process design optimization. However, the applications of surrogate models in the field of wastewater treatment modeling have not been extensively explored in the literature. In this work we present a systematic methodology for construction of powerful surrogate models to be used for the global sensitivity analysis of Benchmark Simulation Model 1 (BSM1) plant. A quasi-random design of experiments technique, Sobol sampling, is employed to generate an experimental design, which is further used to build surrogate models. A class of advanced metamodeling algorithms such as sparse polynomial chaos expansion (PCE) using least angle regression, Kriging interpolation, polynomial chaos Kriging (PCK), radial basis function (RBF) interpolation, multivariate adaptive regression splines (MARS), and a multilayer perceptron type feedforward neural network (ANN) are applied. The generalization error of the developed models has been estimated using holdout cross-validation with coefficient of determination (R2) and root mean squared error (RMSE) being used for evaluation of model predictive accuracy and model selection, respectively. The framework was further investigated forits suitability in a Monte Carlo based global sensitivity analysis using Sobol' method.The results obtained suggest that by following the framework, ANN and Kriging type surrogate models can effectively be constructed and used to estimate Sobol' sensitivity indices of WWTP design parameters.

AB - Surrogate modeling (also referred to as metamodeling) has attracted increased attention from researchers in various fields of engineering due to its use in computationally expensive engineering tasks such as Monte Carlo based global sensitivity analysis and process design optimization. However, the applications of surrogate models in the field of wastewater treatment modeling have not been extensively explored in the literature. In this work we present a systematic methodology for construction of powerful surrogate models to be used for the global sensitivity analysis of Benchmark Simulation Model 1 (BSM1) plant. A quasi-random design of experiments technique, Sobol sampling, is employed to generate an experimental design, which is further used to build surrogate models. A class of advanced metamodeling algorithms such as sparse polynomial chaos expansion (PCE) using least angle regression, Kriging interpolation, polynomial chaos Kriging (PCK), radial basis function (RBF) interpolation, multivariate adaptive regression splines (MARS), and a multilayer perceptron type feedforward neural network (ANN) are applied. The generalization error of the developed models has been estimated using holdout cross-validation with coefficient of determination (R2) and root mean squared error (RMSE) being used for evaluation of model predictive accuracy and model selection, respectively. The framework was further investigated forits suitability in a Monte Carlo based global sensitivity analysis using Sobol' method.The results obtained suggest that by following the framework, ANN and Kriging type surrogate models can effectively be constructed and used to estimate Sobol' sensitivity indices of WWTP design parameters.

KW - Surrogate modelling

KW - Wastewater treatment modelling

KW - Design of experiments

KW - Metamodeling

KW - Global sensitivity analysis

U2 - 10.1016/B978-0-444-64241-7.50313-X

DO - 10.1016/B978-0-444-64241-7.50313-X

M3 - Article in proceedings

VL - 44

SP - 1909

EP - 1914

BT - Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018

A2 - Eden, Mario R.

A2 - Ierapetritou, Marianthi G.

A2 - Towler, Gavin P.

PB - Elsevier

ER -

Al R, Behera CR, Zubov A, Sin G. Systematic framework development for the construction of surrogate models for wastewater treatment plants. In Eden MR, Ierapetritou MG, Towler GP, editors, Proceedings of the 13th International Symposium on Process Systems Engineering – PSE 2018. Vol. 44. Elsevier. 2018. p. 1909-1914. (Computer Aided Chemical Engineering). https://doi.org/10.1016/B978-0-444-64241-7.50313-X