A list-based simulated annealing algorithm with crossover operator for the traveling salesman problem

被引:0
作者
İlhan İlhan
Gazi Gökmen
机构
[1] Necmettin Erbakan University,Department of Mechatronic Engineering, Faculty of Engineering
[2] Necmettin Erbakan University,Department of Educational Sciences, Institute of Educational Sciences
来源
Neural Computing and Applications | 2022年 / 34卷
关键词
Cooling schedule; Genetic edge recombination crossover; Order crossover; Simulated annealing; The Taguchi method; The traveling salesman problem;
D O I
暂无
中图分类号
学科分类号
摘要
The traveling salesman problem (TSP) is one of the most popular combinatorial optimization problems today. It is a problem that is easy to identify but hard to solve. Therefore, it belongs to the class of NP-hard optimization problems, and it is a problem of high time complexity. The TSP can be used to solve various real-world problems. Therefore, researchers use it as a standard test bench for performance evaluation of new algorithms. In this study, a new simulated annealing algorithm with crossover operator was proposed, and it was called LBSA-CO. The LBSA-CO is a population-based metaheuristic method. In this method, a list-based temperature cooling schedule, which can adapt to the topology of the solution space of the problem, was used. The solutions in the population were improved with the inversion, insertion and 2-opt local search operators. The order crossover (OX1) and genetic edge recombination crossover (ER) operators were applied to the improved solutions to accelerate the convergence. In addition, the Taguchi method was used to tune the parameters of the LBSA-CO. The proposed method was tested on 65 well-known TSP instances. The results indicated that this method performs better than the state-of-the-art methods on many instances.
引用
收藏
页码:7627 / 7652
页数:25
相关论文
共 78 条
[1]  
Held M(1962)A dynamic programming approach to sequencing problems J Soc Ind Appl Math 10 196-210
[2]  
Karp RM(1966)Branch-and-bound methods: A survey Oper Res 14 699-719
[3]  
Lawler EL(1991)A branch-and-cut algorithm for the resolution of large-scale symmetric traveling salesman problems SIAM Rev 33 60-100
[4]  
Wood DE(1978)Using cutting planes to solve the symmetric travelling salesman problem Math Program 15 177-188
[5]  
Padberg M(1954)Solution of a large-scale traveling-salesman problem J Oper Res Soc Am 2 393-410
[6]  
Rinaldi G(1973)An effective heuristic algorithm for the traveling-salesman problem Oper Res 21 498-516
[7]  
Miliotis P(2009)General k-opt submoves for the Lin-Kernighan TSP heuristic Math Program Comput 1 119-163
[8]  
Dantzig G(2003)Tour merging via branch-decomposition INFORMS J Comput 15 233-248
[9]  
Fulkerson R(2013)A powerful genetic algorithm using edge assembly crossover for the traveling salesman problem INFORMS J Comput 25 346-363
[10]  
Johnson S(2018)An improved genetic algorithm crossover operator for traveling salesman problem Turkish J Math Comput Sci 9 1-13