交叉概率对改进遗传算法运算速度影响的测试与分析

• 作 者: 弭宝福 徐梅 文士发 何梦莹 孙鲁宁
• 机 构: 东北农业大学工程学院,江西农业大学高等教育研究所,
• 出 处: 数学的实践与认识
• 发表日期: 2014年12期
• 关键词: 遗传算法 交叉概率 运算速度 测试

在现有文献基础上,通过理论分析,提出了交叉概率应取1.通过四个测试函数的测试结果,表明交叉概率越大,运算速度越快,平均运行代数越少,验证了通过理论分析提出的交叉概率应取1的科学性和可行性.同时克服了交叉概率难以确定以及确定交叉概率时缺少理论依据的不足.

On the basis of the existing literature, we propose that the crossover probability should be taken as 1. Based on the test results of the four test functions, it is shown that the larger the crossover probability is, the faster the operation speed is and the smaller the average running algebra is, which verifies that by theoretical analysis The proposed crossover probability should be scientifically and practically feasible and overcome the shortcomings of the lack of theoretical basis for determining the crossover probability and determining the crossover probability.