Original scientific paper

An improved real hybrid genetic algorithm

Weidong Ji ; Department of Computer Science and Information Engineering, Harbin Normal University, Harbin 150025, P. R. China
Jianhua Wang ; Department of Computer Science and Information Engineering, Harbin Normal University, Harbin 150025, P. R. China
Jun Zhang ; Department of Computer Science and Information Engineering, Harbin Normal University, Harbin 150025, P. R. China

Abstract

Aiming at the problem of premature convergence of genetic algorithm and particle swarm algorithm, in order to let the two methods converge to the global optimal solution with the greatest probability and improve the efficiency of the algorithm, the paper will combine improved genetic algorithm with improved particle swarm optimization method to constitute mixed improved algorithm. Through multiple benchmark function used to test the performance of real hybrid genetic algorithm, the results show that hybrid algorithm has good global search ability, fast convergence, good quality of the solution, and good robust performance of its optimization results.

Keywords

genetic algorithm; particle swarm optimization; hybrid algorithm

Hrčak ID:

129044

URI

https://hrcak.srce.hr/129044

Publication date:

29.10.2014.

Article data in other languages: croatian

Visits: 1.910 *