This paper studies the optimal configuration model of non- repairable spare parts whose demand obeys non- Poisson distribution to overcome the limitation of existing models. The characteristics of non-repairable spare parts in two-echelon inventory system were analyzed, and the assumption of the demand obeying Poisson distribution was relaxed, so that a universal model for the satisfaction rate of non-repairable spare parts whose demand obeys non-Poisson distribution was established. On the basis of that, an optimal configuration model for non-repairable spare parts in the two-echelon inventory system was established by setting minimum configuration cost as the optimization target and satisfaction rate as the constraint. The optimization scheme was solved by Genetic Algorithm. The optimization and verification results show that the satisfaction rate of the system reached 95.02% with minimum total cost, which is in accordance with results calculated by Opus 10, demonstrating the feasibility and reliability of the model.

non-repairable spare parts  /  optimal configuration  /  non-Poisson distribution  /  Genetic Algorithm