摘要
将基于实数编码的遗传算法应用于平面度的评定。根据尺寸和公差的数学定义 ,建立完全符合最小区域条件的平面度评定的数学模型 ,并在此基础上给出遗传算法的适应度函数。随后详细地介绍了算法的实现步骤 ,在基于实数编码的基础上 ,遗传选择操作采用一种正比选择策略———转轮法 ,遗传交叉操作采用简单算术交叉法 ,而遗传变异操作是随机均匀实数变异操作。最后对文献 [5 ]的实验数据进行了评定 ,仿真结果表明该算法不仅合理 ,而且效率高、精度高 ,优于其它算法。
A real_coded genetic algorithm is proposed to solve the form evaluation of planes.According to the mathematical definition of dimensional and tolerancing principles,a mathematical model of flatness evaluation in accordance with the minimum zone condition is presented and the fitness function of genetic algorithm is given.Then the implementation steps of the algorithm are dwelt on. Based on the real_coded method,the genetic operation is composed of roulette wheel selection,real_valued simple crossover and uniform mutation.Finally the data in literature are verified by the proposed approach.The results show that the algorithm is superior to others in terms of efficiency and quality of solution in the field of dimension and tolerance evaluation.
出处
《光学精密工程》
EI
CAS
CSCD
2002年第1期36-41,共6页
Optics and Precision Engineering
基金
国家自然科学基金资助项目 (No .5 975 0 2 5 )
关键词
平面度测量
实数编码
遗传算法
flatness measurement
real_coding
genetic algorithm
