An improved decoding procedure and seeker optimization algorithm for reverse logistics network design problem


SUBULAN K., BAYKASOĞLU A., Saltabas A.

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS, vol.27, no.6, pp.2703-2714, 2014 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 27 Issue: 6
  • Publication Date: 2014
  • Doi Number: 10.3233/ifs-141335
  • Journal Name: JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.2703-2714
  • Keywords: Reverse logistics network design, seeker optimization algorithm, particle swarm optimization, fuzzy reasoning, SUPPLY CHAIN NETWORK, INTEGER PROGRAMMING-MODEL, GENETIC ALGORITHM
  • Dokuz Eylül University Affiliated: Yes

Abstract

Recently, Reverse Logistics (RL) and product recovery options such as recycling, remanufacturing and reusing have become important issues due to the environmental, economical issues and legal regulations. Due to this fact, companies should take into account the utilized recovery option while preparing their strategic planning activities (like network design) instead of using traditional production planning models. However, since RL network design problems are in the class of NP-hard, solving large scaled problems by exact algorithms is very difficult. Therefore, many meta-heuristics optimization algorithms have been proposed to provide near optimal solutions for supply chain, RL and closed-loop supply chain network design problems in the literature. In this paper, available decoding algorithms for solving generic RL design problems are revised so as to balance the problem without introducing any dummy node on the chromosome. Moreover, the proposed decoding procedure takes into account "equal transportation cost" situation. Then, a priority-based seeker optimization algorithm (SOA) which utilizes fuzzy reasoning procedure is developed for solution of the problem. In order to test performance of the algorithm, a numerical example is provided and obtained results are compared with particle swarm optimization (PSO) algorithm which is another swarm intelligence technique. Computational results show that SOA is superior to PSO in terms of both solution quality and computational time for the examined RL network design problem.