Optimization of adiabatic styrene reactor: A hybrid multiobjective differential evolution (H-MODE) approach

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

In this study, a hybrid strategy of multiobjective differential evolution (hybrid-MODE) algorithm is proposed that consists of an evolutionary algorithm for global search and a deterministic algorithm for local search. To begin, the proposed algorithm is tested for its performance using a benchmark test function (KUR) as case study 1 with the nondominated sorting genetic algorithm-II (NSGA-II) (both binary- and real-coded versions), strength Pareto evolutionary algorithm (SPEA), Pareto archived evolutionary strategy (PAES), multiobjective differential evolution (MODE), and an improved strategy of MODE algorithms. Subsequently, the multiobjective optimization of an industrial adiabatic styrene reactor is carried out as case study 2, employing a prevalidated model using the hybrid-MODE algorithm and an improved strategy of MODE. Four cases (three sets of two-objective optimization, cases 1-3, and one set of three-objective optimization, case 4) are considered consisting of simultaneous maximization of styrene productivity, selectivity, and yield with four decision variables and two constraints. The proposed algorithm converges to a better set of nondominated solutions (possibly a Pareto front) as compared to the nondominated solutions obtained using NSGA and an improved strategy of MODE algorithms. The limitations of previous studies are reported in terms of key decision variables. The hybrid strategy of MODE is found to converge to the true Pareto front more rapidly (in fewer function evaluations), resulting in a well-diversified Pareto front as compared to the stand-alone evolutionary approach.

Original languageEnglish
Pages (from-to)11115-11132
Number of pages18
JournalIndustrial and Engineering Chemistry Research
Volume48
Issue number24
DOIs
Publication statusPublished - 2009

Fingerprint

Styrene
Evolutionary algorithms
Function evaluation
Multiobjective optimization
Sorting
Genetic algorithms
Productivity

ASJC Scopus subject areas

  • Chemical Engineering(all)
  • Chemistry(all)
  • Industrial and Manufacturing Engineering

Cite this

@article{03948f1e09fa465890c54b3c3633cb17,
title = "Optimization of adiabatic styrene reactor: A hybrid multiobjective differential evolution (H-MODE) approach",
abstract = "In this study, a hybrid strategy of multiobjective differential evolution (hybrid-MODE) algorithm is proposed that consists of an evolutionary algorithm for global search and a deterministic algorithm for local search. To begin, the proposed algorithm is tested for its performance using a benchmark test function (KUR) as case study 1 with the nondominated sorting genetic algorithm-II (NSGA-II) (both binary- and real-coded versions), strength Pareto evolutionary algorithm (SPEA), Pareto archived evolutionary strategy (PAES), multiobjective differential evolution (MODE), and an improved strategy of MODE algorithms. Subsequently, the multiobjective optimization of an industrial adiabatic styrene reactor is carried out as case study 2, employing a prevalidated model using the hybrid-MODE algorithm and an improved strategy of MODE. Four cases (three sets of two-objective optimization, cases 1-3, and one set of three-objective optimization, case 4) are considered consisting of simultaneous maximization of styrene productivity, selectivity, and yield with four decision variables and two constraints. The proposed algorithm converges to a better set of nondominated solutions (possibly a Pareto front) as compared to the nondominated solutions obtained using NSGA and an improved strategy of MODE algorithms. The limitations of previous studies are reported in terms of key decision variables. The hybrid strategy of MODE is found to converge to the true Pareto front more rapidly (in fewer function evaluations), resulting in a well-diversified Pareto front as compared to the stand-alone evolutionary approach.",
author = "Gujarathi, {Ashish M.} and Babu, {B. V.}",
year = "2009",
doi = "10.1021/ie901074k",
language = "English",
volume = "48",
pages = "11115--11132",
journal = "Industrial & Engineering Chemistry Product Research and Development",
issn = "0888-5885",
publisher = "American Chemical Society",
number = "24",

}

TY - JOUR

T1 - Optimization of adiabatic styrene reactor

T2 - A hybrid multiobjective differential evolution (H-MODE) approach

AU - Gujarathi, Ashish M.

AU - Babu, B. V.

PY - 2009

Y1 - 2009

N2 - In this study, a hybrid strategy of multiobjective differential evolution (hybrid-MODE) algorithm is proposed that consists of an evolutionary algorithm for global search and a deterministic algorithm for local search. To begin, the proposed algorithm is tested for its performance using a benchmark test function (KUR) as case study 1 with the nondominated sorting genetic algorithm-II (NSGA-II) (both binary- and real-coded versions), strength Pareto evolutionary algorithm (SPEA), Pareto archived evolutionary strategy (PAES), multiobjective differential evolution (MODE), and an improved strategy of MODE algorithms. Subsequently, the multiobjective optimization of an industrial adiabatic styrene reactor is carried out as case study 2, employing a prevalidated model using the hybrid-MODE algorithm and an improved strategy of MODE. Four cases (three sets of two-objective optimization, cases 1-3, and one set of three-objective optimization, case 4) are considered consisting of simultaneous maximization of styrene productivity, selectivity, and yield with four decision variables and two constraints. The proposed algorithm converges to a better set of nondominated solutions (possibly a Pareto front) as compared to the nondominated solutions obtained using NSGA and an improved strategy of MODE algorithms. The limitations of previous studies are reported in terms of key decision variables. The hybrid strategy of MODE is found to converge to the true Pareto front more rapidly (in fewer function evaluations), resulting in a well-diversified Pareto front as compared to the stand-alone evolutionary approach.

AB - In this study, a hybrid strategy of multiobjective differential evolution (hybrid-MODE) algorithm is proposed that consists of an evolutionary algorithm for global search and a deterministic algorithm for local search. To begin, the proposed algorithm is tested for its performance using a benchmark test function (KUR) as case study 1 with the nondominated sorting genetic algorithm-II (NSGA-II) (both binary- and real-coded versions), strength Pareto evolutionary algorithm (SPEA), Pareto archived evolutionary strategy (PAES), multiobjective differential evolution (MODE), and an improved strategy of MODE algorithms. Subsequently, the multiobjective optimization of an industrial adiabatic styrene reactor is carried out as case study 2, employing a prevalidated model using the hybrid-MODE algorithm and an improved strategy of MODE. Four cases (three sets of two-objective optimization, cases 1-3, and one set of three-objective optimization, case 4) are considered consisting of simultaneous maximization of styrene productivity, selectivity, and yield with four decision variables and two constraints. The proposed algorithm converges to a better set of nondominated solutions (possibly a Pareto front) as compared to the nondominated solutions obtained using NSGA and an improved strategy of MODE algorithms. The limitations of previous studies are reported in terms of key decision variables. The hybrid strategy of MODE is found to converge to the true Pareto front more rapidly (in fewer function evaluations), resulting in a well-diversified Pareto front as compared to the stand-alone evolutionary approach.

UR - http://www.scopus.com/inward/record.url?scp=73349142316&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=73349142316&partnerID=8YFLogxK

U2 - 10.1021/ie901074k

DO - 10.1021/ie901074k

M3 - Article

AN - SCOPUS:73349142316

VL - 48

SP - 11115

EP - 11132

JO - Industrial & Engineering Chemistry Product Research and Development

JF - Industrial & Engineering Chemistry Product Research and Development

SN - 0888-5885

IS - 24

ER -