A novel small-population genetic algorithm based on adaptive mutation and population entropy sampling

The application of Interactive Evolutionary Computation (IEC) requires that corresponding evolutionary algorithms should still have effective and stable performance with small population. The corresponding study of genetic algorithms as evolutionary algorithm is analyzed. A novel mutation strategy based on population entropy sampling to adjust the population diversity intentionally and adaptively is designed, and a new adaptive genetic algorithm with small population is proposed. The proposed algorithm integrating roulette wheel selection and one-point crossover can avoid the premature convergence more effectively and obtain more precise global optimal solutions with fast convergence speed, which makes the proposed algorithm suitable for the application of IEC. Seven multimodal benchmark functions are used to test the performance of the proposed algorithm and the results show that the new algorithm is more effective and stable.