Effects of demand variations on inventory management has long been recognized in academics and practice. Disney et al. (2015) showed that the periodic review order-up-to policy is more feasible for controlling fill rate and avoid over investment of warehouse under auto-correlated normally distributed demand of consumer electronics. Jalali and Nieuwenhuyse (2015) concluded that simulation method is more feasible in reflecting real stock keeping and managing process in terms of the composition of the supply chain, commodity types, stock picking and delivery, and lead-time variations. They also indicated that simulation optimization method is better than analytical models in investigation of real world inventory management problems, which involve with intractable random demands and multiple operational periods in supply chains with multiple echelons.