Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach

Ali Shafiei, Mohammad Ali Ahmadi, Seyed Hayan Zaheri, Alireza Baghban, Ali Amirfakhrian, Reza Soleimani

Research output: Contribution to journalArticle

52 Citations (Scopus)

Abstract

For the design and development of new processes of gas sweetening using ionic liquids (ILs), as promising candidates for amine solutions, an amazing model to predict the solubility of acid gases is of great importance. In this direction, in the current study, the capability of artificial neural networks (ANNs) trained with back propagation (BP) and particle swarm optimization (PSO), to correlate the solubility of H2S in 11different ILs have been investigated. Different structures of three-layer feed forward neural network using acentric factor (ω), critical temperature (Tc), critical pressure (Pc) of ILs accompanied by pressure (P) and temperature (T), as input parameters, were examined and an optimized architecture has been proposed as 5-9-1.Implementation of these models for 465 experimental data points collected from the literature shows coefficient of determination (R2) of 0.99218 and mean squared error (MSE) of 0.00025 from experimental values for PSO-ANN predicted solubilities while the values of R2 = 0.95151 and MSE = 0.00335 were obtained for BP-ANN model. Therefore, through PSO training algorithm we are able to attain significantly better results than with BP training procedure based on the statistical criteria.

Original languageEnglish
Pages (from-to)525-534
Number of pages10
JournalJournal of Supercritical Fluids
Volume95
DOIs
Publication statusPublished - Jan 1 2014
Externally publishedYes

Fingerprint

Ionic Liquids
Hydrogen Sulfide
machine learning
hydrogen sulfide
Hydrogen sulfide
Backpropagation
Ionic liquids
Particle swarm optimization (PSO)
Learning systems
estimating
solubility
Solubility
Neural networks
liquids
Gases
optimization
education
Feedforward neural networks
Amines
critical pressure

Keywords

  • Artificial neural network
  • Hydrogen sulfide
  • Ionic liquids
  • Particle swarm optimization
  • Prediction
  • Solubility

ASJC Scopus subject areas

  • Chemical Engineering(all)
  • Condensed Matter Physics
  • Physical and Theoretical Chemistry

Cite this

Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach. / Shafiei, Ali; Ahmadi, Mohammad Ali; Zaheri, Seyed Hayan; Baghban, Alireza; Amirfakhrian, Ali; Soleimani, Reza.

In: Journal of Supercritical Fluids, Vol. 95, 01.01.2014, p. 525-534.

Research output: Contribution to journalArticle

Shafiei, Ali ; Ahmadi, Mohammad Ali ; Zaheri, Seyed Hayan ; Baghban, Alireza ; Amirfakhrian, Ali ; Soleimani, Reza. / Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach. In: Journal of Supercritical Fluids. 2014 ; Vol. 95. pp. 525-534.
@article{46c720910ac243b9b99bc41edb6333fc,
title = "Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach",
abstract = "For the design and development of new processes of gas sweetening using ionic liquids (ILs), as promising candidates for amine solutions, an amazing model to predict the solubility of acid gases is of great importance. In this direction, in the current study, the capability of artificial neural networks (ANNs) trained with back propagation (BP) and particle swarm optimization (PSO), to correlate the solubility of H2S in 11different ILs have been investigated. Different structures of three-layer feed forward neural network using acentric factor (ω), critical temperature (Tc), critical pressure (Pc) of ILs accompanied by pressure (P) and temperature (T), as input parameters, were examined and an optimized architecture has been proposed as 5-9-1.Implementation of these models for 465 experimental data points collected from the literature shows coefficient of determination (R2) of 0.99218 and mean squared error (MSE) of 0.00025 from experimental values for PSO-ANN predicted solubilities while the values of R2 = 0.95151 and MSE = 0.00335 were obtained for BP-ANN model. Therefore, through PSO training algorithm we are able to attain significantly better results than with BP training procedure based on the statistical criteria.",
keywords = "Artificial neural network, Hydrogen sulfide, Ionic liquids, Particle swarm optimization, Prediction, Solubility",
author = "Ali Shafiei and Ahmadi, {Mohammad Ali} and Zaheri, {Seyed Hayan} and Alireza Baghban and Ali Amirfakhrian and Reza Soleimani",
year = "2014",
month = "1",
day = "1",
doi = "10.1016/j.supflu.2014.08.011",
language = "English",
volume = "95",
pages = "525--534",
journal = "Journal of Supercritical Fluids",
issn = "0896-8446",
publisher = "Elsevier",

}

TY - JOUR

T1 - Estimating hydrogen sulfide solubility in ionic liquids using a machine learning approach

AU - Shafiei, Ali

AU - Ahmadi, Mohammad Ali

AU - Zaheri, Seyed Hayan

AU - Baghban, Alireza

AU - Amirfakhrian, Ali

AU - Soleimani, Reza

PY - 2014/1/1

Y1 - 2014/1/1

N2 - For the design and development of new processes of gas sweetening using ionic liquids (ILs), as promising candidates for amine solutions, an amazing model to predict the solubility of acid gases is of great importance. In this direction, in the current study, the capability of artificial neural networks (ANNs) trained with back propagation (BP) and particle swarm optimization (PSO), to correlate the solubility of H2S in 11different ILs have been investigated. Different structures of three-layer feed forward neural network using acentric factor (ω), critical temperature (Tc), critical pressure (Pc) of ILs accompanied by pressure (P) and temperature (T), as input parameters, were examined and an optimized architecture has been proposed as 5-9-1.Implementation of these models for 465 experimental data points collected from the literature shows coefficient of determination (R2) of 0.99218 and mean squared error (MSE) of 0.00025 from experimental values for PSO-ANN predicted solubilities while the values of R2 = 0.95151 and MSE = 0.00335 were obtained for BP-ANN model. Therefore, through PSO training algorithm we are able to attain significantly better results than with BP training procedure based on the statistical criteria.

AB - For the design and development of new processes of gas sweetening using ionic liquids (ILs), as promising candidates for amine solutions, an amazing model to predict the solubility of acid gases is of great importance. In this direction, in the current study, the capability of artificial neural networks (ANNs) trained with back propagation (BP) and particle swarm optimization (PSO), to correlate the solubility of H2S in 11different ILs have been investigated. Different structures of three-layer feed forward neural network using acentric factor (ω), critical temperature (Tc), critical pressure (Pc) of ILs accompanied by pressure (P) and temperature (T), as input parameters, were examined and an optimized architecture has been proposed as 5-9-1.Implementation of these models for 465 experimental data points collected from the literature shows coefficient of determination (R2) of 0.99218 and mean squared error (MSE) of 0.00025 from experimental values for PSO-ANN predicted solubilities while the values of R2 = 0.95151 and MSE = 0.00335 were obtained for BP-ANN model. Therefore, through PSO training algorithm we are able to attain significantly better results than with BP training procedure based on the statistical criteria.

KW - Artificial neural network

KW - Hydrogen sulfide

KW - Ionic liquids

KW - Particle swarm optimization

KW - Prediction

KW - Solubility

UR - http://www.scopus.com/inward/record.url?scp=84916641476&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84916641476&partnerID=8YFLogxK

U2 - 10.1016/j.supflu.2014.08.011

DO - 10.1016/j.supflu.2014.08.011

M3 - Article

VL - 95

SP - 525

EP - 534

JO - Journal of Supercritical Fluids

JF - Journal of Supercritical Fluids

SN - 0896-8446

ER -