基于分级策略的演化算法

为了避免演化算法过早收敛，保持种群多样性，增加算法的搜索能力，本文提出基于分级策略的演化算法．即通过对种群进行分级，来度量种群的多样性，衡量算法是否陷入局部最优，协调种群多样性和精英策略之间的矛盾，再根据种群分布的多样性设计一种有效的半一致交叉算子与单重均匀变异算子。

Evolutionary Algorithm Based on Hierarchical Strategy

In order to avoid the evolutionary algorithms convergence prematurely,maintain population diversity, increase the local search ability of the algorithm, This paper proposed a method based on hierarchical evolutionary algorithm. Through to the population classification, to measure the diversity of population, to measure the algorithm whether into local optimum, Coordination of population diversity and the contradiction between the elite strategy, and design of an efficient semi uniform crossover operator and mutation operator of single heavy.