Improved Heuristics for Multi-Stage Requirements Planning Systems

Abstract

Most of the recent studies of heuristic lot-sizing techniques for multi-stage material requirements planning systems have investigated the problem in the context of a single stage. In this paper, the multi-stage problem is first modeled analytically to indicate the potential errors inherent in the commonly proposed single-pass, stage-by-stage approaches (e.g., Wagner-Whitin). Then, based on this analysis, several simple cost modifications are suggested to improve the global optimality of these procedures. The effectiveness of these alternative modifications is tested through a series of simulation experiments. It is concluded that some of these modifications consistently improve the performance of the stage-by-stage approaches.