In this paper, we study the reoptimization problems which arise when a new node is added to an optimal solution of a traveling salesman problem (TSP) instance or when a node is removed. We show that both reoptimization problems are NP-hard. Moreover, we show that, while the cheapest insertion heuristic has a tight worst-case ratio equal to 2 when applied to a TSP instance, it guarantees, in linear time, a tight worst-case ratio equal to 3/2 when used to add the new node and that also the simplest heuristic to remove a node from the optimal tour guarantees a tight ratio equal to 3/2 in constant time. © 2003 Wiley Periodicals, Inc.
[1]
Leen Stougie,et al.
Algorithms for the On-Line Travelling Salesman1
,
2001,
Algorithmica.
[2]
Markus W. Schäffter,et al.
Scheduling with Forbidden Sets
,
1997,
Discret. Appl. Math..
[3]
Nicos Christofides.
Worst-Case Analysis of a New Heuristic for the Travelling Salesman Problem
,
1976,
Operations Research Forum.
[4]
David S. Johnson,et al.
Computers and Intractability: A Guide to the Theory of NP-Completeness
,
1978
.
[5]
Gerhard Reinelt,et al.
Traveling Salesman Problem
,
2012
.