Improved Fitness Proportionate Selection-Based Genetic Algorithm
Yu Fengrui,Fu Xueliang,Li Honghui,Dong Gaifang
DOI: https://doi.org/10.2991/icmit-16.2016.26
2016-01-01
Abstract:Genetic algorithms are typical swarm intelligence techniques based on the mechanics of natural selection and natural genetics, which combines artificial survival of the fittest concept with genetic operations abstracted from nature. Since the genetic algorithm has good global search capability, as well as the parallel nature of other advantages, it has been widely used in combinatorial optimization, machine learning, signal processing field, adaptive control and artificial life and so on. It is one of the key technologies related to modern intelligent calculation. Fitness proportionate selection, as a common selection method for GA, is usually implemented with method of roulette wheel selection. In this paper, An improved selection method based on fitness proportionate selection was presented. Computational results show that the method which proposed in this paper improved the result precision and better astringency by solving TSP problem.