224 European Journal of Operational Research 41 (1989) 224-231
North-Holland
Theory and Methodology
Algorithms to solve the orienteering
A comparison *
problem:
C. Peter KELLER
Department of Geography, University of Victoria, Victoria, B.C., Canada
Abstract: The multiobjective vending problem (MVP) (Keller, 1985) is a generalization of the traveling
salesman problem (TSP) where it is not necessary to visit all nodes in the problem definition. A special
case of the MVP is the orienteering problem (OP) (Tsiligirides, 1984; Golden et al., 1985). The problems
are NP-hard. A heuristic was designed to solve the general MVP (Keller, 1985). Three other heuristics were
designed to solve the OP (Tsiligirides, 1984; Golden et al., 1985). The algorithms underlying these
heuristics are outlined. Their performances are compared against three OP problems.
(
Keywords: Traveling salesman problem, generalization, orienteering problem, heuristics, performance
comparison
1. Introduction
The multiobjective vending problem (MVP)
(Keller, 1985) is a generalization of the traveling
salesman problem (TSP) where it is not necessary
to visit all nodes in the problem definition. The
objective is to examine the trade-off relationship
between maximizing reward potential by visiting
as many nodes as possible, at the same time
minimizing penalty for traveling the links by visit-
ing as few nodes as possible. A conceptually simi-
lar problem concerns the orienteering problem
(OP) discussed by Tsiligirides (1984), and Golden
et al. (1985). It has been demonstrated that both
problems are NP-hard (Garey and Johnson, 1979;
Keller, 1985; Golden et al., 1985).
The objective of this paper is to evaluate and
discuss the performance of four algorithms written
to solve the above two problems for three sets of
OP data. Tsiligirides (1984), Keller (1985) and
Goldert et al. (1985) have suggested that the MVP
and the OP have numerous applications. The rela-
tive performance of algorithms to solve these
problems are therefore of interest. The paper com-
mences by outlining the two problem definitions.
Next, the different solution approaches are sum-
marized. The performances of the algorithms are
compared using Tsiligirides (1984) three score
orienteering problems.
2. Problem definition
* This research has been supported by NSERC grant no. A
6533.
Received November1987; revised May 1988
The TSP is conceptually defined as follows:
Given a set of n nodes, determine the shortest
complete circuit that connects all nodes so that
0377-2217/89/$3.50 © 1989, ElsevierSciencePublishers B.V. (North-Holland)