Abstract According to the phenomenon that fresh agricultural products retailers always sell more products in the morning and the late afternoon while between them they sell less, a single sale period is divided into three parts. This paper considers customers arrive following the Nonhomogeneous Poisson Process with a random parameter, and each takes a quantity depending on the products’ freshness and price. It introduce the Risk Preference Coefficient to describe retailers’ risk preference degree, focus on how freshness decline coefficient and risk preference coefficient affect retailer’s ordering decision, and provide an optimal ordering policy for each of them. Conclusion shows an ordering policy considering retailer’s risk preference and deteriorating rate of product, is more efficient for agricultural products. Finally, the conclusion is drawn by a numerical example.
2012, Vol. 9 
