Research on a novel genetic algorithm based on adaptive evolution in dual population
Yan Tai-shan, Guanqi Guo, Li Wu · 2011
Considering the limitation of standard genetic algorithm such as premature convergence and low convergence speed, an improved genetic algorithm based on adaptive evolution in dual population (DPAGA) is proposed. In this algorithm, the new population produced by selecting operation is regarded as the main population. The population composed by the individuals washed out by selecting operation is regarded as the subordinate population. The individual evolution strategy in the main population is different from that in the subordinate population. The crossover operators and mutation operators are all adjusted non-linearly and adaptively. Experiments are taken on 8 typical testing functions. The experimental results show that this algorithm is stable and fast. It is valid in solving optimization problems.