Evaluation of a new type of sequential lot sizing techniques

The classical dynamic lot size problem without backlogging is usually solved with the aid of various heuristics. Most heuristics are sequential, i.e. the demand is considered period for period, and a decision whether to include the demand in a certain period in the preceding batch is taken without regarding the future demand. Recently, it has been shown how to design a sequential lot sizing rule that will optimize the average performance, provided that a typical demand looks like a sequence of independent and identically distributed random numbers. The purpose of this paper is to evaluate if and how this methodology can be implemented in practice. The new lot sizing techniques are evaluated in a simulation study for different types of demand.