Replenishment decision support system based on modified particle swarm optimization in a VMI supply chain

This paper proposes a replenishment decision support system, based on response surface methodology (RSM) and modified traveling particle swarm optimization (TPSO). Cross-border cooperation lengthens the distance for transportation and in turn widens the disparity in inventory cost among business partners, thus the accentuated importance of inventory management. Therefore, this paper solves a two-stage stochastic dynamic lot-sizing problem with two-phased transportation cost under a vendor managed inventory. Modified TPSO is proposed to solve a sub-problem, the nonlinear mixed integer programming. In this algorithm, which is executed with a perturbation policy, each move is a feasible solution. RSM is being used to determine the optimal replenishment condition. The result of the experiment indicates that the proposed approach lowers the cost for both the buyer and the vendor. Moreover, the solution quality using modified TPSO was tested and compared with that of the Solver tool in Excel and 3 lot-sizing decision rules