基于有导向变异算子的GM-EA算法

GM-EA:guided mutation evolutionary algorithm

为了进一步提高演化算法的效率,提出基于有导向变异算子的GM-EA算法(guided mutation evolutio-nary algorithm)。通过结合粒子群优化的方法改进郭涛算法,更好地利用当前最优解指导变异,并将算法分为探索与开采两个阶段;在开采阶段基于模拟退火方法决定是否用新个体取代旧个体,在巩固所获取的建筑块成分的同时,尽可能克服早熟收敛问题。实验结果证明了新算法的有效性。

To design a more effective evolutionary algorithm,this paper introduced a new guided mutation evolutionary algorithm by combining Guotao algorithm with the idea from particle swarm optimization,which focused on exploiting the global best solution in population to direct the mutation.In order to preserve the components of building-blocks and avoid the premature problem,separated the search process as the exploration phase and exploitation phase,and in exploitation phase simulated annealing was applied as the replace policy.The experimental results show that the proposed algorithm is significantly superior to Guotao algorithm.