A mixed integer robust programming model for two-echelon inventory routing problem of perishable products

Abstract

To solve the inventory routing problem of perishable products with time window constraints, a mixed integer linear programming (MILP) model is constructed to minimize the total cost. Due to the uncertainty of market demand, the MILP model is further transformed into mixed integer robust programming (MIRP) model by introducing uncertain sets (box, ellipsoid and polyhedron). Experiments with actual data show that as demand uncertainty increases, although MIRP models will pay corresponding robust costs, they can achieve better robustness. In addition, the comparison shows that the ellipsoid set MIRP model can achieve a higher level of logistics and distribution services.