Stocking Decision for Spare Parts in A System with General Demand Distribution

In order to investigate the stocking decision for a two-echelon inventory system of non-repairable spare parts with multiple demand distributions, we present a universal model to calculate non-repairable spare parts fill rate. Because of the diversity of non-repairable spare parts, the assumption of general demand distribution is made instead of Poisson demand distribution. The two-echelon inventory system consists of a warehouse and several locations. To obtain a high fill rate level of the system in limit cost, an optimization stocking decision model of the system is established by setting total cost as optimization target and systematic fill rate as constraints. The optimal stocking scheme is solved by Genetic Algorithm. Normal distribution and logistic distribution are taken for numerical examples to demonstrate the performance of universal model. The result shows that fill rate of system can be 95.02% with the minimum total cost. Sensitivity analyses give the relationship between parameters and fill rate. The research result can offer the decision basis for military supply chain.