We consider the problem of determining the optimal reorder intervals R and order-up-to levels S in a multi-echelon supply chain system where all echelons are assumed to have fixed ordering costs and to operate with a (R, S) policy with stationary nested power-of-two reorder intervals. By using the guaranteed service approach to model the multi-echelon system facing a stochastic demand, we formulate the problem as a deterministic optimisation model in order to simultaneously determine the optimal R and S parameters as well as the guaranteed service times. The model is a non-linear integer programming (NLIP) problem with a non-convex and non-concave objective function including rational and square root terms. Then, we propose a sequential optimisation procedure (SOP) to obtain near-optimal solutions with reasonable computational time. The numerical study demonstrates that for a general acyclic multi-echelon system with randomly generated parameters, the SOP is able to obtain near-optimal solutions of about 0.46% optimality gap in average in a few seconds. Moreover, we propose an improved direct approach using a global optimiser, bounding the decision variables in the NLIP model and considering the SOP solution as an initial solution. Numerical examples illustrate that this reduces significantly the computational time.Keywords: inventory control; multi-echelon system; guaranteed service model; power-of-two policies
IntroductionMany real-world supply chains are complex multi-echelon systems consisting of suppliers, manufacturers, wholesalers and retailers that have geographically dispersed facilities. One challenge these supply chains face is the efficient management of inventory when demand is uncertain, operating costs are important and customer service requirements are high. This requires specifying the inventory policy at different echelons so that to minimise the total cost of the whole multi-echelon system subject to customer service levels (Simchi-Levi and Zhao 2012). The guaranteed service model (GSM) which is among the relevant approaches that can be used in multi-echelon inventory systems has gained interest in recent years. In particular, this model enabled to realise important benefits in practice in general multi-echelon systems which combine distribution and assembly systems (see e.g. Billington et al. 2004;Farasyn et al. 2011).In this paper, we build on the power-of-two (PO2) and the GSM research to find a reasonable solution to the problem of simultaneously optimising the reorder intervals and order-up-to levels for general multi-echelon systems facing stochastic demand. Finding an optimal policy for this problem would be extremely difficult. Indeed, the optimal policy is not known even for two-echelon distribution systems with deterministic demand (Snyder and Shen 2011). In order to deal with demand variations, we use the original assumptions of the GSM that are the guaranteed service time and the bounded demand assumptions. Besides, we assume that each stage of the supply chain operates with a peri...