Following the theoretical studies of J.B. Robinson and H.W. Kuhn in the late 1940s and the early 1950s, G.B. Dantzig, R. Fulkerson, and S.M. Johnson demonstrated in 1954 that large instances of the TSP could be solved by linear programming. Their approach remains the only known tool for solving TSP instances with more than several hundred cities; over the years, it has evolved further through the work of M. Grötschel, S. Hong, M. Junger, P. Miliotis, D. Naddef, M. Padberg, W.R. Pulleyblank, G. Reinelt, G. Rinaldi, and others. We enumerate some of its refinements that led to the solution of a 13,509-city instance.

Dantzig, Fulkerson, and Johnson (1954) introduced the cutting-plane method as a means of attacking the traveling salesman problem; this method has been applied to broad classes of problems in combinatorial optimization and integer programming. In this paper we discuss an implementation of Dantzig et al.'s method that is suitable for TSP instances having 1,000,000 or more cities. Our aim is to use the study of the TSP as a step towards understanding the applicability and limits of the general cutting-plane method in large-scale applications.

The first computer implementation of the Dantzig-Fulkerson-Johnson cutting-plane method for solving the traveling salesman problem, written by Martin, used subtour inequalities as well as cutting planes of Gomory’s type. The practice of looking for and using cuts that match prescribed templates in conjunction with Gomory cuts was continued in computer codes of Miliotis, Land, and Fleischmann. Grötschel, Padberg, and Hong advocated a different policy, where the template paradigm is the only source of cuts; furthermore, they argued for drawing the templates exclusively from the set of linear inequalities that induce facets of the TSP polytope. These policies were adopted in the work of Crowder and Padberg, in the work of Grötschel and Holland, and in the work of Padberg and Rinaldi; their computer codes produced the most impressive computational TSP successes of the nineteen eighties. Eventually, the template paradigm became the standard frame of reference for cutting planes in the TSP. The purpose of this paper is to describe a technique

Many classes of valid and facet-inducing inequalities are known for the family of polytopes associated with the Symmetric Travelling Salesman Problem (STSP), including subtour elimination, 2-matching and comb inequalities. For a given class of inequalities, an exact separation algorithm is a procedure which, given an LP relaxation vector x∗ , nds one or more inequalities in the class which are violated by x , or proves that none exist. Such algorithms are at the core of the highly successful branch-and-cut algorithms for the STSP. However, whereas polynomial time exact separation algorithms are known for subtour elimination and 2-matching inequalities, the complexity of comb separation is unknown. A partial answer to the comb problem is provided in this paper. We de ne a generalization of comb inequalities and show that the associated separation problem can be solved efficiently when the subgraph induced by the edges with x ∗ e ¿0 is planar. The separation algorithm runs in O(n³) time, where n is the number of vertices in the graph.

We consider most of the known classes of valid inequalities for the graphical travelling salesman polyhedron and compute the worst-case improvement resulting from their addition to the subtour polyhedron. For example, we show that the comb inequalities cannot improve the subtour bound by a factor greater than ~. The corresponding factor for the class of clique tree inequalities is 8, while it is 4 for the path configuration inequalities. Keywords: Polyhedral combinatorics; Valid inequalities; Travelling salesman; Worst-case analysis 1.

The symmetric Generalized Travelling Salesman Problem (GTSP) is a variant of the classical symmetric Travelling Salesman Problem, in which the nodes are partitioned into clusters and the salesman has to visit at least one node for each cluster. A different version of the problem, called E-GTSP, arises when exactly one node for each cluster has to be visited. Both GTSP and E-GTSP are NP-hard problems, and find practical applications in routing and scheduling. In this paper we model GTSP and E-GTSP as integer linear programs, and study the facial structure of the corresponding polytopes. In a companion paper (Fischetti, Salazar and Toth [5]), the results described in this work have been used to design a branch-and-cut algorithm for the exact solution of instances up to 442 nodes.

Given an undirected graph G = (V; E) and a cost vector c 2 IR E , the weighted girth problem is to find a circuit in G having minimum total cost. This problem is in general NP-hard since the traveling salesman problem can be reduced to it. A promising approach to hard combinatorial optimization problems is given by the so-called cutting plane methods. These involve linear programming techniques based on a partial description of the convex hull of the incidence vectors of possible solutions. We consider the weighted girth problem in the case where G is the complete graph K n and study the facial structure of the circuit polytope P n C and some related polyhedra. In the appendix we give complete characterizations of P n C for n up to 6.

A convex polytope can either be described as convex hull of vertices or as solution set of a finite number of linear inequalities and equations. Whereas both representations are equivalent from a theoretical point of view, they are not when optimization problems over the polytope have to be solved. Moreover, it is a challenging task in practical computation to convert one description into the other. In this paper we address the efficient computation of the facet structure of polytopes given by their vertices and present new computational results for polytopes which are of interest in combinatorial optimization. Keywords: polytope, convex hull, combinatorial optimization 1 Introduction Hard combinatorial optimization problems are often attacked with branch-and-cut methods. These methods strongly rely on knowledge about the structure of the polytope that is defined as convex hull of the 0/1 incidence vectors of feasible solutions. In particular, knowledge about linear equations and ineq...

Abstract not found

This paper is the second in a series of two papers dedicated to the separation problem in the symmetric traveling salesman polytope. The first one gave the basic ideas behind the separation procedures and applied them to the separation of Comb inequalities. We here address the problem of separating inequalities which are all, in a way or another, a generalization of Comb inequalities. These are namely Clique Trees, Path, Ladder inequalities. Computational results are reported for the solution of instances of the TSPLib using the branch and cut framework ABACUS.