New neighborhoods and an iterated local search algorithm for the generalized traveling salesman problem

For a given graph with vertex set that is partitioned into clusters, the generalized traveling salesman problem (GTSP) of finding cost-minimal cycle contains exactly one every cluster. We introduce three new GTSP neighborhoods allow simultaneous permutation sequence clusters and selection vertices from each The some known literature are combined an effective iterated local search (ILS) for GTSP. ILS performs straightforward random neighborhood within applies ordinary record-to-record acceptance criterion. computational experiments on four symmetric standard libraries show that, purposeful refinements, can compete state-of-the-art algorithms. • Effective problem. Local uses several in variable descent (VND). Three traditional neighborhoods. Random VND compared prioritization.