On the Recoverable Traveling Salesman Problem

Abstract

In this paper we consider the Recoverable Traveling Salesman Problem (TSP). Here the task is to find two tours simultaneously, such that the intersection between the tours is at least a given minimum size, while the sum of travel distances with respect to two different distance metrics is minimized. Building upon the classic double-tree method, we derive a 4-approximation algorithm for the Recoverable TSP. We also show that if the required size of the intersection between the tours is constant, a 2-approximation guarantee can be achieved, even if more than two tours need to be constructed. We discuss consequences for approximability results in the more general area of recoverable robust optimization.