一种算子操作概率的自适应策略

An Adaptive Strategy for Operator Probability

在分析传统遗传算法易发生早熟收敛的基础上,本文提出了一种改进算法IGA。IGA采用一种新的交叉率变异率自适应变化的策略,防止在进化初期种群不发生变化造成算法早熟。数值计算结果表明,利用改进策略的IGA搜索整体性较强,可以有效地避免早熟现象产生,提高全局优化能力。

This paper analyses tersely the reasons resulting in prematurity arising in the practice of genetic algorithms （GA） and puts forth an improved algorithm IGA. IGA adopts the strategy of adaptive cross rate and mutation rate to solve the prematurity problem. The simulation results show that IGA has the characteristics of high convergence and can effectively avoid prematurity to improve global optimization.