Scanning-area algorithms for clustered TSP

Abstract

One of the base problems of combinatorial optimization, that has wide applications, is the Traveling Salesman Problem (TSP). Rapid growth of its size (thousands of points) is the main feature now, and many extensions of this problem with specific properties are studied: dynamic vehicle routing problems; systems with call (the first-aid, the courier's, the taxi), systems with time windows, and systems of delivery. They need development of the special algorithms, which provide high-quality solutions in reasonable time. That is why subsequent perfection of existing algorithms and development of new ones is relevant. In this paper, a set of strategies were proposed and evaluated, in order to solve specific TSP problems in which the points are under the form of clusters.