Simulation And Optimization For Two-Echelon Inventory System With Complicated Stochastic Demand And Random Leadtime

A computer simulation model is built for a two-echelon inventory system with complicated stochstic demand and random leadtime, with the reorder point being confined by the cycle service level. A optimal solution to minimize the total cost can be obtained from this model.As it is difficult to determine the parameters in the muti-period inventory policies using the analytical methods, simulation software Flexsim and statistical analysis software ExpertFit are applied to establish the simulation models of the inventory system with the demand following complicated or joint distributions. The desired parameters of the inventory policy can be obtained by running the simulation models if given a customer service level. As the established query tables with the demand following simple distributions, query tables with the demand following complicated or joint distributions are given in a certain range.And then a simulation model of a two-echelon inventory system with stochastic variables is built with every inventory unit using (R, Q) policy. The processes to get the solution are determined based on the theory of two-echelon inventory. The amount of work can be reduced greatly with the confined reorder point by the cycle service level and the model operability for the optimization gets better. The key of the simulation optimization is to get the demand in leadtime for every inventory unit. The mean of out of stock amount needed in total cost calculation can get from the random numbers of demand in leadtime or from the distribution function that the demand in leadtime follows. The distribution function of the order delay time that obtained from the simulation model is not normal definitely.