| This thesis focuses on some inventory management policies for substitutable and perishable items under demand uncertainty. A set of perishable products with fixed shelf lives is considered under an (R,Si) system of inventory control where demand for a preferred product can be satisfied by a substitute product with a known probability, in the event of a stockout of the preferred product. While taking demand substitution and product expiration into account, the retailer is faced with the decision of determining the order-up-to level, Si, for each product i which maximizes expected total profit, given a common review period, R, determined exogenously.Under demand uncertainty, the problem detailed in this thesis involves stochastic optimization. An exact closed form expression, however, for expected profits becomes difficult for certain parameter values involving product shelf-life, product substitution, and lead time. As an alternative approach, order replenishment, demand consumption, substitution, and product expiration can be effectively modeled using discrete-event simulation. Through a discrete-event simulation model, each realization of the profit function can be evaluated for a selected value of Si, and a mean profit value can be estimated after a number of replications of a simulation run. In order to find the best Si solution, the technique of simulation-optimization is used. |
