Abstract: Due to difficulties in accurately predicting not only the timing but also and magnitude of disasters, which leads to the serious expiration and shortage problems that exist in perishable emergency materials, the strategies of return near shelf life and urgent supply are introduced into the perishable inventory system to reduce the expiration and shortage losses. Referring the classical newsboy inventory problem, two stochastic models for return and urgent supply strategies with uncertainties in demand time and demand quantity are constructed and verified in this paper. In our models, the opportunity losses are considered, and the performances of both strategies in reduced risks and added values are analyzed. In order to describe the opportunity losses accurately, the unit cost is adopted per unit time. With further efforts to measure the performances consisting of reduced risk and added values, the likelihoods of expiration and shortage as risks and differences of opportunity losses as strategies' values are calculated. The analytical outcomes indicate that the strategies of return near shelf life and urgent supply are not always favorable, the implementations of these two strategies should satisfy certain conditions. Thus, the sufficient advantage conditions for both strategiesare given in this paper, in addition, a standard furtherly to balance the reduced risks and added values is also set in our study helping decision-makerselect better inventory policies. For an intuitive perspective, the parameters in our perishable emergency inventory system are estimated by collecting real disaster data online and relevant literatures, and some numerical examples are computed to presented sensitivities of some key parameters and impacts of different distributed functions. From the numerical results, it is concluded that the predefined return time is more robust in return model than return-supply model and the added values does not always increase with the increase of return time; distributed functions of demand time and demand quantity have obvious effects on added values; the effects of uncertainties degree are not uniform and monotonic. The paper makes a preliminary exploration for return and urgent supply strategies in perishable inventory system, which may drive more future works for sustainable perishable inventory networks.

Key words: perishable emergency materials, return close-to-expiry, emergency supply, stochastic demand time, added value