Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management

Tamás A. Varga, András A. Király, J. Abonyi

Research output: Chapter in Book/Report/Conference proceedingChapter

4 Citations (Scopus)

Abstract

Advanced inventory management in complex supply chains requires effective and robust nonlinear optimization due to the stochastic nature of supply and demand variations. Application of estimated gradients can boost up the convergence of Particle Swarm Optimization (PSO) algorithm but classical gradient calculation cannot be applied to stochastic and uncertain systems. In these situations Monte-Carlo (MC) simulation can be applied to determine the gradient. We developed a memory-based algorithm where instead of generating and evaluating new simulated samples the stored and shared former function evaluations of the particles are sampled to estimate the gradients by local weighted least squares regression. The performance of the resulted regional gradient-based PSO is verified by several benchmark problems and in a complex application example where optimal reorder points of a supply chain are determined. © 2013

Original languageEnglish
Title of host publicationSwarm Intelligence and Bio-Inspired Computation
PublisherElsevier Inc.
Pages403-422
Number of pages20
ISBN (Print)9780124051638
DOIs
Publication statusPublished - 2013

Fingerprint

Particle swarm optimization (PSO)
Supply chains
Data storage equipment
Function evaluation
Uncertain systems
Stochastic systems
Monte Carlo simulation

Keywords

  • Gradient
  • Inventory management system
  • Monte-Carlo simulation
  • Particle swarm optimization

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Varga, T. A., Király, A. A., & Abonyi, J. (2013). Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management. In Swarm Intelligence and Bio-Inspired Computation (pp. 403-422). Elsevier Inc.. https://doi.org/10.1016/B978-0-12-405163-8.00019-3

Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management. / Varga, Tamás A.; Király, András A.; Abonyi, J.

Swarm Intelligence and Bio-Inspired Computation. Elsevier Inc., 2013. p. 403-422.

Research output: Chapter in Book/Report/Conference proceedingChapter

Varga, TA, Király, AA & Abonyi, J 2013, Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management. in Swarm Intelligence and Bio-Inspired Computation. Elsevier Inc., pp. 403-422. https://doi.org/10.1016/B978-0-12-405163-8.00019-3
Varga TA, Király AA, Abonyi J. Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management. In Swarm Intelligence and Bio-Inspired Computation. Elsevier Inc. 2013. p. 403-422 https://doi.org/10.1016/B978-0-12-405163-8.00019-3
Varga, Tamás A. ; Király, András A. ; Abonyi, J. / Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management. Swarm Intelligence and Bio-Inspired Computation. Elsevier Inc., 2013. pp. 403-422
@inbook{def73beab1f4464ab286ac90b17ca231,
title = "Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management",
abstract = "Advanced inventory management in complex supply chains requires effective and robust nonlinear optimization due to the stochastic nature of supply and demand variations. Application of estimated gradients can boost up the convergence of Particle Swarm Optimization (PSO) algorithm but classical gradient calculation cannot be applied to stochastic and uncertain systems. In these situations Monte-Carlo (MC) simulation can be applied to determine the gradient. We developed a memory-based algorithm where instead of generating and evaluating new simulated samples the stored and shared former function evaluations of the particles are sampled to estimate the gradients by local weighted least squares regression. The performance of the resulted regional gradient-based PSO is verified by several benchmark problems and in a complex application example where optimal reorder points of a supply chain are determined. {\circledC} 2013",
keywords = "Gradient, Inventory management system, Monte-Carlo simulation, Particle swarm optimization",
author = "Varga, {Tam{\'a}s A.} and Kir{\'a}ly, {Andr{\'a}s A.} and J. Abonyi",
year = "2013",
doi = "10.1016/B978-0-12-405163-8.00019-3",
language = "English",
isbn = "9780124051638",
pages = "403--422",
booktitle = "Swarm Intelligence and Bio-Inspired Computation",
publisher = "Elsevier Inc.",

}

TY - CHAP

T1 - Improvement of PSO Algorithm by Memory-Based Gradient Search-Application in Inventory Management

AU - Varga, Tamás A.

AU - Király, András A.

AU - Abonyi, J.

PY - 2013

Y1 - 2013

N2 - Advanced inventory management in complex supply chains requires effective and robust nonlinear optimization due to the stochastic nature of supply and demand variations. Application of estimated gradients can boost up the convergence of Particle Swarm Optimization (PSO) algorithm but classical gradient calculation cannot be applied to stochastic and uncertain systems. In these situations Monte-Carlo (MC) simulation can be applied to determine the gradient. We developed a memory-based algorithm where instead of generating and evaluating new simulated samples the stored and shared former function evaluations of the particles are sampled to estimate the gradients by local weighted least squares regression. The performance of the resulted regional gradient-based PSO is verified by several benchmark problems and in a complex application example where optimal reorder points of a supply chain are determined. © 2013

AB - Advanced inventory management in complex supply chains requires effective and robust nonlinear optimization due to the stochastic nature of supply and demand variations. Application of estimated gradients can boost up the convergence of Particle Swarm Optimization (PSO) algorithm but classical gradient calculation cannot be applied to stochastic and uncertain systems. In these situations Monte-Carlo (MC) simulation can be applied to determine the gradient. We developed a memory-based algorithm where instead of generating and evaluating new simulated samples the stored and shared former function evaluations of the particles are sampled to estimate the gradients by local weighted least squares regression. The performance of the resulted regional gradient-based PSO is verified by several benchmark problems and in a complex application example where optimal reorder points of a supply chain are determined. © 2013

KW - Gradient

KW - Inventory management system

KW - Monte-Carlo simulation

KW - Particle swarm optimization

UR - http://www.scopus.com/inward/record.url?scp=84883950645&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883950645&partnerID=8YFLogxK

U2 - 10.1016/B978-0-12-405163-8.00019-3

DO - 10.1016/B978-0-12-405163-8.00019-3

M3 - Chapter

AN - SCOPUS:84883950645

SN - 9780124051638

SP - 403

EP - 422

BT - Swarm Intelligence and Bio-Inspired Computation

PB - Elsevier Inc.

ER -