Simulation-based optimization method for three-echelon network inventory system of a supply chain

A single-item, three-echelon network inventory system of a supply chain is considered that consists of two manufactures, two distribution centers and three retailers and implements the reorder point, order quantity (r, Q) inventory policy. The policy decision problem is formulated in order to minimize the total cost of the system subject to customer arrival Possion process, stochastic demand, production capacity and other uncertain factors. Because most of such complex and stochastic problems are not simple to apply analytic method only and simulation is not an optimization tool, a simulation-based optimization method consisting of a simulation model and particle swarm optimization algorithm is developed to obtain the optimal (r, Q) policies. A numerical experiment is given to illustrate the simulation-based optimization method efficient.