文献类型：期刊文献

中文题名：基于多种群实数编码遗传算法的圆度误差评定

英文题名：Evaluation of Roundness Error Based on RMGA

作者：缪立栋[1];郭慧[1];张帆[1]

机构：[1]华东理工大学机械与动力工程学院,上海200237

年份：2010

卷号：31

期号：2

起止页码：43

中文期刊名：工程图学学报

外文期刊名：Journal of Engineering Graphics

收录：北大核心:【北大核心2008】；CSCD:【CSCD_E2011_2012】；

语种：中文

中文关键词：计算机应用;圆度误差评定;多种群遗传算法;圆度

外文关键词：computer application; roundness error evaluation; multi-population genetic algorithm; circularity;

摘要：针对实数编码的遗传算法(RGA)在圆度误差评定中存在的早熟收敛、精度较差及运算速度慢等缺点,分析了多种群遗传算法的模型结构,利用多种群遗传算法的并行性,将其应用于圆度误差评定。实验证明该方法有效地提高了算法的精度和收敛速度,能够快速评定圆度误差。
In order to improve the problems of real-coding based genetic algorithm(RGA),such as premature convergence,rough result and low speed of calculation,the model structure of multi-population genetic algorithm(RMGA) is analyzed,and the parallel characteristics of this algorithm is applied to the roundness error evaluation.Experiments show that this algorithm can improve the accuracy and convergence in assessment of roundness error effectively.