Optimal partly repairable spare inventory policy by MDP under warranty

Aiming at the problem of spares inventory in warranty logistics, this paper applies Markov modeling to the optimization of spares during the warranty period which considers the case where the repairable item is discarded, analyzes the transfer process of spares in operation, and stock, repair and discard of spares. In the Markov model, the actual situation of installation quantity of components in the system is finite, is taken into consideration and an optimal strategy is introduced. By applying a convex optimization algorithm based on the Markov model, the optimization spares configuration scheme is determined to minimize the cost under a specified availability. Meanwhile this paper builds an application case for this model, and verifies the model and algorithm proposed in this paper by simulation. There are three innovation points of this research. 1) A repairable item that may be discarded is taken into consideration in the warranty period. 2) Build partly repairable spares optimization model by Markov under variational warranty period. 3) A suitable convex optimization algorithm, which can find the optimal point through solving a large number of iterative calculations quickly, is designed.