The Diversity Of Cellular Genetic Algorithm Research

Cellular genetic algorithm is the main cellular automata theory as the foundation. Based on cellular space structure, genetic, crossover and mutation performed in the neighborhood between genetic algorithms. Evolutionary algorithm has been the biggest problem is the existence of how to control the balance of Exploration/Exploitation balance. For the same cellular genetic algorithm, maintaining population diversity and Exploration/Exploitation balance in the evolutionary process is the focus of the study.There are a variety of cellular spatial structures in cellular genetic algorithm. By changing the cellular spatial structure will be able to improve the population diversity of the evolutionary process. Cellular genetic algorithm is able to maintain population diversity in cellular space neighborhood structure of configuration parameters in the evolutionary process. Population diversity measures has been an important research direction for evolution algorithm, genetic algorithms has been proposed to a number of population diversity measures, mainly phenotype, standard deviation, entropy type and genotype in the study. This four measurement methods can effectively measure the true diversity of the population in the evolutionary process, although the genotype is the most complex kind of measurement, but it reflects the diversity of the population is the most realistic through feasibility studies. This paper studies the population individual measures from the point of genotype with genotype coded divided into binary coded, decimal coded and permutation coded. And now domestic and international research about the information of cellular genetic algorithm diversity is extremely limited, articles against cellular genetic algorithm performance in-depth study on the basis of relevant literature data. Emphasis analyzes the methods to maintaining genotype population diversity for cellular genetic algorithm, as well as factors discussed in the initial population diversity.This article mainly aims at the cellular genetic algorithm during evolution there will often be Exploration/Exploitation balance problem, expounds the importance and significance of maintaining diversity. Cellular genetic algorithms as an important part in the evolutionary algorithm, its principle, performance has a greater advantages in maintaining the diversity of the population. This paper analyzes in detail the basic influence factors of population diversity and measurement methods, and the diversity of individual genotypes studied, combining cellular space structure model, the genotypic diversity the measurement method has entered a new level, and summarizes several methods of measurement is more reliable, at the same time performance for cellular genetic algorithm insufficiency, proposed the improvement algorithm, has improved the ability of maintaining the iterative process of population diversity, finally based on improved numerical simulation results on the start of expansion, expansion discussion on the diversity of initial population, puts forward the roulette method, truncation selection, stakes selection and other select methods.