Adaptation ethnic group evolution algorithm

Enlightened by the analysis method of modern anthropology for sociogroup, the ethnic group evolution algorithm (EGEA)*, whose population is formed of several ethnic groups, is proposed. The ethnic group is a kind of clustering form for individuals, and represents the possible evolution trends, which are implicit in population. In EGEA, ethnic groups are used as the competition and generation unit of individuals, which improve the parallelism and diversity of population evolution. In this paper, an improving EGEA, adaptation ethnic group evolution algorithm (AEGEA), is developed. In AEGEA, a series of adaptation control strategies are used to adjust evolution parameters dynamically so as to make the searching process more steady and efficient. The simulations tests of numerical optimization show AEGEA can restrain premature convergence phenomenon effectively during the evolutionary process while increasing the convergence speed greatly.