AN INTEGRATION OF SIMULATED ANNEALING AND GENETIC ALGORITHM FOR TRAVELLING SALESMAN PROBLEM

Aria, Muhammad (2011) AN INTEGRATION OF SIMULATED ANNEALING AND GENETIC ALGORITHM FOR TRAVELLING SALESMAN PROBLEM. Majalah Ilmiah UNIKOM, Volume. ISSN 1411-9374

[img] UNSPECIFIED
volume-81-artikel-12.pdf - Published Version

Download (0B)
Official URL: http://jurnal.unikom.ac.id/jurnal/an-integration-o...

Abstract

This paper presents a new algorithm based on integrating simulated annealing and genetic algorithm to solve travelling salesman problem. We called the pro-posed architecture as The Annealing-Genetic Algorithm (AG). The core of the AG is based on genetic algorithms. Simulated annealing method is used to acceler-ate the convergence of the genetic algorithm by applying the simulated anneal-ing test for all the population members. In order to evaluate the AG, we applies the proposed architecture to Travelling Salesman Problem especially for the larges data problem. The algorithm has been implemented in LabVIEW and tested on several sets of TSP data. For the largest data instance used is 1002 cities (i.e., pr10022), the actual population size used is only 8. For small sets of data (less than 100 cities), AG can find the optimal solutions. These experiment results prove that AG improves the distance 2.25% over SA and 3.09% over GA. But AG computation time is more complex than both SA and GA algorithm, so AG takes 6.47 times slower than SA and 1.3 times than GA algorithm. In appli-cation, AG needs 13.88 bytes units of memory, while SA only needs 10.67 bytes unit memory and GA algorithm needs 31.33 bytes units memory.

Item Type: Article
Uncontrolled Keywords: crossover, genetic algorithm, simulated annealing, travelling salesman problem
Subjects: Jurnal Tercetak > Majalah Ilmiah UNIKOM
Divisions: Universitas Komputer Indonesia > Direktorat > Lembaga Penelitian dan Pengabdian Masyarakat
Depositing User: M.Kom Taryana Suryana
Date Deposited: 30 Nov 2016 09:33
Last Modified: 30 Nov 2016 09:33
URI: https://repository.unikom.ac.id/id/eprint/30511

Actions (login required)

View Item View Item