A Non-Convex Robust Simulation Optimization Model for Inventory Management Problem by System Dynamics

Abstract

Perishable product inventory management is a challenging issue because of its direct effect on companies' profits. The dependence of a product order cost on the order quantity is one of the practical but less examined assumptions in this problem literature. Hence, this paper considers the dependency between the order cost and order quantity as well as between the holding cost and the inventory level. This problem will have a non-convex object, and is not solvable through the usual mathematical methods. Thus, simulation-optimization approach is used to determine the perishable product inventory management policy with stochastic demand. The system dynamics approaches have been used to simulate the problem by minimizing the cost function. The casual diagram, inputs, output, and relation of the system are determined. A numerical example of a hypermarket is presented, and the optimal amount of the objective function is determined with optimization of the input variables via the experimental design's method. Then, to rule out the effects of different errors, a robust optimization of the model is presented. The results show that the proposed replenishment policy could benefit the necessary decisions regarding inventory management and control of the perishable products which count in different errors.