Using Iterated Local Search for Efficiently Packing Unequal Disks in a Larger Circle

Packing unequal disks in a container as small as possible without mutual overlap is a NP-hard problem with a plenty of real world applications. In this paper, we introduced iterated local search to tackle with this problem, and using Critical Element Guided Perturbation (CEGP) strategy to jump out the local minimal, and using BFGS method to get each neighborhood optimized in local search procedure. Experiments showed its efficiency by breaking several world records of a set of benchmark.