Discrete Applied Mathematics 123 (2002) 75-102. doi:10.1016/S0166-218X(01)00338-92016-03-04T18:46:49

EditorialThe traveling salesman problem This Special Issue of Discrete Optimization is devoted to the traveling salesman problem (TSP), its generalizations and modifications. The TSP is a very well studied optimization problem and there have been numerous publications on the TSP and its modifications, including two edited books (the latter was published in 2002 with G. Gutin and A.P. Punnen as editors). The topic continues to attract the attention of researchers from various areas including mathematics and computer science. Generally, TSP publications are scattered among various journals in mathematics, computer science, operations research, artificial intelligence and other related areas. The aim of this Special Issue is to bring to the attention of the reader a selected sample of research papers on the topic. We briefly describe below various papers included in this collection.It is well known that an asymmetric TSP on n vertices could be formulated as a symmetric TSP on 2n vertices. However, the paper by E. Balas, R. Carr, M. Fischetti and N. Simonetti included in this issue is the first article to demonstrate that the transformation can be used to obtain new facets for the symmetric TS polytope from asymmetric TS polytope facets. investigate a new class of polynomial length formulations for the asymmetric TSP. They show that a relaxation of their formulation is tighter than the formulation based on the exponential number of Danzig-Fulkerson-Johnson subtour elimination constraints. They also provide the results of computational experiments showing the efficiency of the proposed formulations.V. Mak and N. Boland study a recently defined generalization of the asymmetric TSP in which all arcs are partitioned into two classes (ordinary and replenishment arcs) and every vertex has a positive weight. A tour is feasible if it uses either type of arc before the total weight of vertices exceeds a certain limit, in which case only a replenishment arc can be used. The authors show that two classes of inequalities, under certain conditions, are facet-defining.A. Grigoriev and J. van de Klundert introduce another generalization of the TSP in which every vertex can be visited several times. The authors investigate how an optimal tour and its value change when the number of visits of each vertex is increased by the same factor.M. Turkensteen, D. Gosh, B. Goldengorin and G. Sierksma discuss an iterative technique for patching within the context of branch-and-bound approaches for the asymmetric TSP. Although already relevant per se, the technique could also be beneficial in other areas such as metaheuristics for the asymmetric TSP.Usually neighborhoods used in local search heuristics are of small polynomial size. This limitation is due to the fact that the best tour in the neighborhood has to be found for the worst case. However, as far back as the 1980s some researchers introduced TSP neighborhoods of exponential size in which the best tour can be computed in polynomial time. The study of exponential neighborhoods for various...

No abstract

No abstract

No abstract