Analytical evaluation of heuristics performance for the single-level lot-sizing problem for products with constant demand

In this paper we examine the performance of several commonly used heuristics for the single-level lot-sizing problem, for products with stationary demand and cost parameters. A particular combination of demand and cost combination, introduced by Blackburn and Millen, called the lot-sizing index (LSI), is shown to be an excellent predictor of heuristics performance for any finite horizon problem. The importance of the LSI as a performance predictor is established through an analysis of the heuristics objective function and the optimal cost function.