Adaptive Hormone Regulation Operator For Genetic Algorithm

摘要

According to the principle of hormone regulation in endocrine system, novel crossover operator and mutation operator for the genetic algorithm were proposed. These adaptive operators replaced the traditional crossover rate and mutation rate by assembling Hill function, and made them self-regulate to maintain the diversity of the population with a reasonable level in the whole procedure of the evolution. Experimental results of four benchmark functions showed that the improved genetic algorithm could not 'only effectively enhance the diversity of the population, but also avoid problems of the premature and local optimum.