摘要
当前统计公差分析与综合中常采用遗传算法对装配件中各零件公差进行优化,以平衡产品成本与合格率的比例.利用遗传算法进行优化虽然可以获得全局最优解,但计算量过大.文中提出利用基于数论网格点集的数论序贯优化算法进行公差优化.与MonteCarlo方法产生的随机点集相比,数论网格点集的分布更为均匀.实例计算结果表明:序贯优化算法在保证获得全局最优解的同时,还显著地降低了计算量。
A sequential algorithm was presented,which is based on the number theory net.It was used to perform the tolerance allocation.The distribution of number theory netis more homogeneous than thatof the point set obtained using Monte Carlo simulation.An example was employed to illustrate the optimal tolerance allocation design,and both efficiency and accuracy of the solution were investigated.This algo- rithm can efficiently obtain the global optimum,and the amount of calculation is considerably reduced to 1 /5 compared to that of genetic algorithm.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2000年第3期374-377,共4页
Journal of Shanghai Jiaotong University
基金
教育部博士点基金!资助项目 (编号 :970 2 483 2 )
关键词
数论
序贯优化算法
公差分析
机械设计
公差综合
numbertheory
sequential numbertheory optimization
numbertheory net
tolerance analysis and synthesis
