Heuristiken zur Ein-Depot-Tourenplanung

München (1995), 120 S.
Buch / Monographie / Fach: Informatik
München, Techn. Univ., Dipl.-Arbeit, 1995
Abstract:
The Vehicle Routing Problem (VRP) is a generalization of the Traveling Salesman Problem (TSP). An amount of goods is assigned to every node of a graph and they are to be collected by vehicles restricted in their capacity and time and brought to a depot. The aim is to minimize the overall costs. The decision problem for vehicle routing is NP-complete, which means that there probably is no polynomial-time algorithm for VRP and we are forced to resort to heuristics. Several heuristics are introduced, including Sweep, several variants of the Savings-procedure, a giant tour algorithm and a $k$-opt algorithm for the improvement of existing tours. The heuristics are tested on graphs representing real-world road networks and are compared in their results and their computational complexity.
Technische Universität München
(in German)