基于种群过早收敛程度定量分析的改进自适应遗传算法

Adaptive Genetic Algorithm to Improve Group Premature Convergence

分析了现有的一些改进算法所提出的评价种群过早收敛程度的指标,讨论了它们的不足;提出了一个概念清楚、运算量小的新指标,并利用该指标给出一种新的交叉概率、变异概率自适应调整策略.仿真实例表明,该方法能及时反映种群在进化过程中的过早收敛程度,不仅能加快计算速度。

Genetic Algorithms (GA) have the disadvantages of premature convergence and low search efficiency. The cross and mutation probability can be modulated to enhance the computation speed and the global convergence efficiency. This is accomplished by introducing an index that reflects the degree of prematurity.