A BACKWARD APPROACH LPC-BASED CONSTRUCTIVE HEURISTIC

ABSTRACT

In this paper, we propose a heuristic for the well-known single-level un-capacitated dynamic lot-sizing problem. The proposed heuristic is based on the average period cost criterion introduced by the heuristic from Silver and Meal [18], also known as Least Period Cost (LPC). Two desirable features of the proposed heuristic are simplicity and the heuristic's ability to complement LPC effectively. We examine the effectiveness of the proposed heuristic by using a set of 20 benchmarking test problems attributed to Baker [2] and Berry [3]. Furthermore, we perform a large-scale simulation involving 80,000 randomly generated problems to evaluate the proposed heuristic against several popular heuristics in the literature. Our computational results show that the proposed heuristic is able to achieve highly competitive results.

Keywords:

• lot-sizing
• heuristics
• material requirements planning
• empirical results