A guided local search based on a fast neighborhood search for the irregular strip packing problem

The irregular strip packing problem asks to place a set of polygons within a rectangular strip of fixed height without overlap, so as to minimize the strip width required. We consider an overlap minimization problem which minimizes the amount of overlap penalty for all pairs of polygons within a given bound of strip width. We propose a local search algorithm which translates a polygon in horizontal and vertical directions iteratively, and incorporate it in metaheuristic approaches called the iterated local search and the guided local search. Computational results show that our algorithm is competitive with other existing algorithms.