【摘要】 本文针对改进遗传算法的搜索性能 ,提出了一种新的变异算子——自适应多位变异算子 ,它根据每一代的每一个体的在当代中的优劣程度 ,自适应地控制其发生变异的位数 .适应度差的个体变异的位数多 ,一方面使差的模式生存机会少 ,另一方面增加了种群的多样性 ,同时也扩大了搜索范围 .我们用模式定理证明了该算子的有效性 ,并且我们作的仿真实验也同样表明该算子能大大改进遗传算法的搜索性能 .更多还原

【Abstract】 This paper proposes a new mutation operator—Adaptive Multi-bit Mutation Operator. It controls the mutation bits of every individual of a certain generation according to this individuals essence adaptively. The bad individuals have more mutation bits than the good. On the one hand it decrease the survival chance of the bad individuals, on the other hand it expand the searching field. We use schemata theorem and our experiments to prove that this operator can improve searching capability of genetic algorithm drastically.更多还原