New model and heuristic solution approach for one-dimensional cutting stock problem with usable leftovers

Abstract

In the one-dimensional cutting stock problem with usable leftovers (1DCSPUL), items of the current order are cut from stock bars to minimize material cost. Here, stock bars include both standard ones bought commercially and old leftovers generated in processing previous orders, and cutting patterns often include new leftovers that are usable in processing subsequent orders. Leftovers of the same length are considered to be of the same type. The number of types of leftovers should be limited to simplify the cutting process and reduce the storage area. This paper presents an integer programming model for the 1DCSPUL with limited leftover types and describes a heuristic algorithm based on a column-generation procedure to solve it. Computational results show that the proposed approach is more effective than several published algorithms in reducing trim loss, especially when the number of types of leftovers is limited.

Keywords:

• cutting and packing
• one-dimensional cutting stock
• column generation
• usable leftovers

Acknowledgments

This research is part of Projects 71371058 and 61363026 supported by the National Natural Science Foundation of China.