Abstract:
A genetic algorithm, based on the classification of the auxiliary group, was proposed. The algorithm overcame the weakness that diversity of the auxiliary group was not perfect, and with prior knowledge it classified the auxiliary group into several subgroups. After the classification, it is more favourable for the auxiliary group and the main group to evolve. Meanwhile, the diversity of the auxiliary group is maintained. The results show that the improved algorithm is more effective than some other existing genetic algorithms. The improved algorithm avoids being trapped into the local optimum, thus deriving the global optimum.

Key words: genetic algorithm; auxiliary population; main population; classification