GrC model in Genetic Algorithm: Artificial Selection Algorithm

Genetic Algorithm (GA), a programming technique that mimics natural evolution as a problem-solving strategy, has become popular since its appearance. It keeps the properties similar to natural selection systems. Many improved GAs has been proposed, however, natural selection essence is not changed. Granular Computing (GrC) brings a new thought for problem solving, it arose a fast growing interests in recent years. This paper build up GrC Model in traditional GA, propose Artificial Selection Algorithm (ASA), which realize intentional reproduction of individuals in a population that have desirable traits. ASA simulates artificial selection behavior of human beings to generate untold diversity in individuals, so as to help guarantee the global convergence and searching efficiency. Artificial selection is a contrast to natural selection, in which the random forces of nature determine becomes the combination of natural selection and artificial selection. The algorithm is validated by benchmark function optimization.