Optimization of Spare Parts Varieties Based on Stochastic DEA Model

Accurate inventory management starts with the scientific and rational classification of numerous varieties of spare parts. This paper presents a stochastic data envelopment analysis (DEA) model to address the problem of optimization of spares varieties under uncertainty. An index system is proposed in terms of product life-cycle process, which contains five design indexes, four operation indexes and five support indexes. Then, the quantification method of the index system is briefly discussed in preparation for mathematical calculation. A stochastic spares optimization model (SSOM) is proposed based on stochastic DEA with the constraints of 14 factors of the index system. The SSOM could be converted into equivalent deterministic models by probability theory, which overcomes the difficulty in solving non-linear programming. A numerical example is given to illustrate the proposed method in terms of ability to provide reasonable inventory management policies.