A New Fuzzy Adaptive Genetic Algorithm Based on Variance and Entropy

This paper proposed a improved genetic algorithm that was fuzzy adaptive genetic algorithm for solving premature convergence. In the new algorithm, it was the use of population variance and entropy to measure diversity of population, and in accordance with the each population of variance and entropy to design fuzzy reasoning system for adaptively controlling crossover probability and mutation probability. Through a multi-function optimization problems simulation, its results prove that this fuzzy adaptive genetic algorithm feasibility and effectiveness.