Optimization of vendor managed inventory of multiproduct EPQ model with multiple constraints using genetic algorithm

• Published in: International journal of advanced manufacturing technology Vol. 71; no. 1-4; pp. 365 - 376
• Main Authors: Pasandideh, Seyed Hamid Reza, Niaki, Seyed Taghi Akhavan, Hemmati Far, Mohammad
• Format: Journal Article
• Language: English
• Published: London Springer London 01-03-2014; Springer Nature B.V
• Subjects: CAE) and Design; Computer-Aided Engineering (CAD; Constraint modelling; Economic models; Engineering; Genetic algorithms; Industrial and Production Engineering; Inventory management; Mechanical Engineering; Media Management; Nonlinear programming; Optimization; Original Article; Storage capacity; Supply chains; Vendor managed inventory; Genetic algorithm; Limited storage; Limited number of orders; Economic production quantity; Multiproduct; Limited budget
• ISSN: 0268-3768; 1433-3015
• DOI: 10.1007/s00170-013-5476-x

The aim of this paper is to investigate the vendor managed inventory (VMI) problem of a single-vendor single-buyer supply chain system, in which the vendor is responsible to manage the buyer’s inventory. To include an extended applicability in real-world environments, the multiproduct economic production quantity model with backordering under three constraints of storage capacity, number of orders, and available budget is considered. The nonlinear programming model of the problem is first developed to determine the near optimal order quantities along with the maximum backorder levels of the products in a cycle such that the total VMI inventory cost of the system is minimized. Then, a genetic algorithm (GA) based heuristic is proposed to solve the model. Numerical examples are given to both demonstrate the applicability of the proposed methodology and to fine tune the GA parameters. At the end, the performance of the proposed GA is compared to the one of the LINGO software using different problem sizes. The results of the comparison study show that, while the solutions do not differ significantly, the proposed GA reaches near optimum solutions in significantly less amount of CPU time.