Search

The Stochastic Guaranteed Service Model with Demand Propagation SGSM-DP is an MILP model that can pro-actively decide on safety stocks and outsourced quantities in a multi-echelon supply chain distribution network in the presence of a discrete demand random distribution and immediate outsourcing options. However, on the one hand, it can only handle small networks, on the other hand its objective does not precisely capture the holding costs for the expected inventory levels. In this paper, new models are presented that overcome both short-comings: The F-SGSM-DP is essentially equivalent to the SGSM-DP but can handle large networks fast. The F-SGSM-DF is an improved model compared to the F-SGSM-DP with a similar scalability but with a more accurate objective. Both new models are based on a representation of a logical material flow with a flow- design in a certain time-expanded network. Numerical results on randomized data confirm that the F-SGSM-DF is faster to compute, predicts better the realized cost, and lead to substantial cost savings in dynamic simulations.