Minimizing average lead time for the coordinated scheduling problem in a two-stage supply chain with multiple customers and multiple manufacturers


Yilmaz Ö. F. , PARDALOS P. M.

COMPUTERS & INDUSTRIAL ENGINEERING, vol.114, pp.244-257, 2017 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 114
  • Publication Date: 2017
  • Doi Number: 10.1016/j.cie.2017.10.018
  • Title of Journal : COMPUTERS & INDUSTRIAL ENGINEERING
  • Page Numbers: pp.244-257
  • Keywords: Two-stage supply chain scheduling, Artificial bee colony algorithm, Simulated annealing algorithm, Retching mechanisms, Average lead time, BEE COLONY ALGORITHM, GENETIC ALGORITHMS, TRANSPORTATION, OPTIMIZATION, DELIVERY

Abstract

In this study, the two-stage supply chain scheduling problem with multiple customers and multiple manufacturers is considered. The first stage consists of m manufacturers (suppliers), while the second stage contains q vehicles, each of which distributes the batches from the manufacturers to the customers. Multiple customers and average lead time are two most important issues in practice; however, no study has been carried out so far to investigate these two issues together for the two-stage supply chain scheduling problem. The main contribution of this study is to coordinate production and distribution decisions to obtain an effective scheduling in a two stage supply chain that contains multiple customers and multiple manufacturers. A mixed integer linear optimization model is developed to formulate the problem with the average lead time objective. Because the problem has been shown-to be NP-hard, a hybrid artificial bee colony and simulated annealing (HABCSA) algorithm is introduced and used to solve the problem. In addition, a lower bound (LB) and several structural properties for the problem are presented and different batching mechanisms are developed based on these properties. For the purpose of performance analysis of HABCSA with different batching mechanisms, detailed computational experiments are conducted using random instances which are generated based on real aluminum production data for different capacity levels. The experimental results indicate that the HABCSA heuristic consistently outperforms the Genetic Algorithm (GA) and the Artificial Bee Colony (ABC) algorithm for each capacity level.