In this paper, we apply control-theoretic approach to the design of inventory policy for systems with perishable goods. In the considered systems, the stock used to fulfill unknown, variable demand is subject to exponential decay. It is replenished from multiple supply sources characterized by different lead times. The challenge is to achieve high demand satisfaction with minimum costs when replenishment orders are realized with non-negligible delay. In contrast to the classical stochastic, or heuristic approaches, a formal design methodology based on discrete time sliding mode (DSM) control is employed. The proposed DSM controller with the sliding plane selected for a dead-beat scheme ensures full demand satisfaction for arbitrary bounded demand. It achieves a given service level with smaller holding costs and reduced order-to-demand variance ratio as compared to the classical order-up-to policy.
Published in:
Automation Science and Engineering, IEEE Transactions on
(Volume:9
,
Issue:
2
)
Date of Publication: April 2012