Analysis of TSP: Simulated Annealing and Genetic Algorithm Approaches

Year 2020, Volume: 6 Issue: 1, 23 - 28, 31.03.2020

Ahmet Reha Botsalı Kemal Alaykıran

https://doi.org/10.22399/ijcesen.637445

Abstract

This paper analyzes the performance of the popular heuristic methods ‘Simulated Annealing (SA)’ and ‘Genetic Algorithm (GA)’  on  the symmetric TSP.  TSP is a well-known combinatorial optimization problem in NP-complete class. NP-completeness of TSP originates many specific approximation algorithms to find optimal or near optimal solutions in a reasonable time. On the other hand, both SA and GA are general purpose heuristic methods that are applicable to almost every kind of problem whose solution lies inside a search space. The performance of SA and GA depends on many factors such as the nature of the problem, design of the algorithm, parameter values, etc. In this paper, a GA and an SA algorithm are given  and their performance with re-spect to several factors is analyzed. The algorithms are tested on some benchmark problems (TSPLIB) which are obtainable via Internet from http://elib.zib.de/pub/mp-testdata/tsp/tsplib/tsp/index.html.

Keywords

TSP, Simulated Annealing (SA), Genetic Algorithms (GA), Integer Programming (MIP)

References

• ILOG Cplex. World Wide Web, https://www.ibm.com/tr-tr/products/ilog-cplex-optimization-studio .
• Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, Clifford Stein 2001. Introduction to Algorithms. The MIT Press Cambridge, Massachusetts.
• TSPLIB. Library of Sample Instances for the TSP. http://elib.zib.de/pub/mp-testdata/tsp/tsplib/tsp/index.html