Solving the economic lot scheduling problem with deteriorating items using genetic algorithms

Ming-Jong Yao*, Jian Xiong Huang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

25 Scopus citations

Abstract

The focus of this study is to solve the Economic Lot Scheduling Problem (ELSP) with deteriorating items using the extended basic period approach under Power-of-Two (PoT) policy. The ELSP is concerned with the lot sizing and scheduling decision of n items, and, PoT policy requires the replenishment frequency of each item to be a PoT integer. Since the conventional ELSP is shown as an NP-hard problem, we propose a hybrid genetic algorithm (HGA) that is equipped with a feasibility testing procedure, namely, Proc FT, and a binary search heuristic to efficiently solve the ELSP with deteriorating items. We use Proc FT to test the feasibility of the local optimum for the set of replenishment frequencies obtained by the evolutionary process of genetic algorithm. When the local optimum for the set of replenishment frequencies is infeasible, we employ the proposed binary search heuristic to search for a feasible solution with the minimal objective value. We use six benchmark examples to evaluate the performance of the proposed hybrid GA. Our numerical results demonstrate that the proposed hybrid GA is an efficient approach in solving the ELSP with deteriorating items. Therefore, we conclude that the proposed HGA could be a favorable decision support tool for managers in the food industry to efficiently derive their production scheduling and lot-sizing strategies for deteriorating items.

Original languageEnglish
Pages (from-to)309-322
Number of pages14
JournalJournal of Food Engineering
Volume70
Issue number3
DOIs
StatePublished - 1 Oct 2005

Keywords

  • Deterioration
  • Genetic algorithm
  • Lot size
  • Scheduling

Fingerprint Dive into the research topics of 'Solving the economic lot scheduling problem with deteriorating items using genetic algorithms'. Together they form a unique fingerprint.

Cite this