A coordinated location-inventory problem with supply disruptions: A two-phase queuing theory–optimization model approach

• Published in: Computers & industrial engineering Vol. 142; p. 106326
• Main Authors: Liu, Yunhua, Dehghani, Ehsan, Jabalameli, Mohammad Saeed, Diabat, Ali, Lu, Chung-Cheng
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.04.2020
• Subjects: Disruption; Facility location; Genetic algorithm; Inventory management; Queuing theory; Supply Chain; Facility location; Inventory management; Queuing theory; Disruption; Genetic algorithm; Supply Chain
• ISSN: 0360-8352; 1879-0550
• DOI: 10.1016/j.cie.2020.106326

•This paper deals with a coordinated location-inventory problem in a stochastic supply.•Facilities are subject to random supply disruptions.•The demand and replenishment lead-time are also considered to be stochastic.•A two-phase approach based on the queuing theory and optimization model techniques is devised. Despite their prominent role in real-life, supply disruptions and inherent variability in parameters have been virtually overlooked in the area of the location-inventory problem. To push forward the relevant literature, addressing the aforementioned issues is of paramount importance. In this sense, this paper deals with a coordinated location-inventory problem in a stochastic supply chain system where facilities are subject to random supply disruptions. The demand and replenishment lead-times are also considered to be hemmed in by uncertainties. To tackle the location-inventory problem in the concerned supply chain, a two-phase approach based on the queuing theory and optimization model techniques is devised. Applying the queuing theory, the first phase tackles the uncertainty issues and gains some performance measures of the system. The attained results are later incorporated into the optimization model. In the second phase, the optimization model determines the strategic and tactical decisions across the supply chain in an integrated manner. Because the presented model is complicated to solve through exact methods, a tailored hybrid genetic algorithm embedded with direct search method is exploited, which is capable of finding quality solutions in an efficient way. Eventually, various sensitivity analyses are conducted from which interesting insights are derived.