Using the Bees Algorithm to solve combinatorial optimisation problems for TSPLIB

Abstract

The Bees Algorithm (BA) is a metaheuristic algorithm to find good solutions to optimisation problems in reasonable computing times. This paper is the first to report on the use of the BA to solve 9 combinatorial optimisation problems (COP) with more than 100 cities from TSPLIB rigorously to test the performance of the algorithm. The work employed a basic version of the BA for COP and TSPLIB datasets involving between 100 and 200 cities. The results obtained show that deviations from the best-found tour lengths for the datasets with 100 cities and 200 cities were approximately 2.5% and 7.5%. The reason for this jump in deviations was that the number of iterations was kept constant for all experiments while the solution space increased factorially with the number of cities. This research can be replicated and modified through Google Colab.