Computer and Modernization ›› 2010, Vol. 1 ›› Issue (11): 12-15.doi: 10.3969/j.issn.1006-2475.2010.11.004

• 算法设计与分析 • Previous Articles     Next Articles

A New Hybrid Genetic Algorithm for Traveling Salesman Problem

HU Zhi-wei1, QIE Pei2, ZHAO Xin-chao3, LI Xian-xu1   

  1. 1.School of Information and Telecommunication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China; 2.School of Computer, Beijing University of Posts and Telecommunications, Beijing 100876, China; 3.Department of Mathematics, School of Science, Beijing University of Posts and Telecommunications, Beijing 100876, China
  • Received:2010-08-10 Revised:1900-01-01 Online:2010-11-25 Published:2010-11-25

Abstract: The paper presents an improved hybrid genetic algorithm to solve the Traveling Salesman Problem (TSP). Based on the traditional genetic algorithm, Guo Tao algorithm is introduced to the crossover operation, which keeps good diversity and global search capability to overcome the shortcomings of premature convergence. PSO algorithm is introduced to the mutation operation, which accelerates the convergence rate, improves the solution quality, and finds the optimal solution more quickly. Verified by many examples from TSPLIB, the obtained results are better than the available optimal solutions, which demonstrate the efficiency and efficacy of the algorithm.

Key words: traveling salesman problem, genetic algorithm, Guo Tao algorithm, Particle Swarm Optimization

CLC Number: