A fast algorithm for two-dimensional pallet loading problems of large size

Abstract

This paper concerns the two-dimensional pallet loading problem (PLP), which requires the determination of the orthogonal layout that loads the maximum number of identical small rectangles (i.e., boxes or products) onto a large rectangle (i.e., pallet or container) without overlapping. Although many algorithms have been developed for this problem, the large amount of time required to find efficient layouts for a large PLP presents great practical difficulties. In this paper, we develop a heuristic that finds efficient layouts with low complexity. We also propose a new algorithm, using the heuristic as a sub-algorithm, which rapidly finds complicated solutions having a 5-block structure. Finally, computational results show that the new algorithm can be successfully applied to large PLPs with sizes exceeding 6800 boxes.