Dynamic allocation of uncertain supply for the perishable commodity supply chain

Dynamic logistics control of a perishable commodity is especially crucial due to the difficulties in preservation of freshness, disposal of deteriorated commodities, and uncertainty of supply caused by seasonal fluctuation or abrupt variation of the weather. First, this paper formally presents the Dynamic Allocation Problem with Uncertain Supply (DAP/US) for the perishable commodity supply chain (PC-SC). The objectives of the DAP/US problem are to maximize the total net profit of the strategic alliance of the PC-SC and to determine the optimal orders placed to suppliers and the resultant amount of perishable commodities allocated to retailers. Secondly, a two-stage extended-Genetic Algorithm (eGA) is developed to control the dynamic orders and allocation quantities to prioritized suppliers and retailers, respectively. Thirdly, simulation experiments are conducted and it is shown that eGA demonstrates promising performance under various sizes of problem domains and different statuses of supply uncertainty. Lastly, anlytical simulations are also conducted to compare eGA with the traditional approach that ignores the differences among suppliers' supply uncertainties. The simulation results show that eGA achieves great improvement in both the net profit and shortage rates for all sizes of the PC-SCs, with half of their suppliers and retailers having high supply capabilities and fast turnover demands, respectively.