带自适应动态变异和二次变异的差分进化算法

Differential evolution with adaptive dynamic mutation & second mutation

为解决差分进化算法在解决多目标优化问题时的多样性与收敛性之间的平衡维持难题,首先提出了一种基于自适应动态变异和非支配解二次变异的改进差分进化算法。该算法的核心是将前N代进化的群体多样性值作为多样性判别准则,根据群体多样性变化情况自适应地选择对应的变异算子产生新个体;其次提出通过对所存档Pareto非支配解进行二次变异来增加新个体解群的优解质量和数量,以同时改进算法的多样性和收敛速度。仿真结果表明,与标准差分进化算法和改进的基于分类排序的Pareto遗传算法相比,所提算法在收敛性、分布性与分散度性能指标上都有较好的表现,多样性和收敛性之间的平衡维持能力则远优于另两种算法。

In order to keep balance between diversity and convergence of Differential Evolution（DE）in solving multiobjective optimization,an improved DE based on adaptive dynamic mutation and second mutation of non-dominance solution was proposed.The core of this algorithm was：a new diversity dominance based on the previous population diversity was presented so that DE could adaptively select the corresponding mutation function for generating new individual according to the diversity variety.In addition,second mutation based on Pareto non-dominance solution archive was executed in order to improve quantity and quality of new individual solutions,algorithm＇s diversity and convergence speed were improved simultaneously.Compared to standard DE and non-dominated sorting genetic algorithm II,simulation results showed that this algorithm had better convergence,diversity and spread,and it was also superior to other two algorithms in keeping balance between diversity and convergence.