基于分类数据的质量控制算法研究.PDF

第 19卷 第 6期 运 筹 与 管 理 Vo.l 19, No. 6 2010年 12月 OPERAT IONS RESEARCH AND MANAGEMENT SCIENCE Dec. 2010 韩 凯, 王凯波 ( , 10008 ) : , , , , , , , , : ; ; ; ; : F22 : A : 2010) P rocess Quality Control U sing Categorical Observation s HAN K ai, WANG Kaibo (D ep artm en t of IndustrialE ng ineering, T singhua Univ ersity, B eij ing 10008 , China ) Ab stract: Categorical observations are common and easily captured. Much of current w ork focuses on process control and quality mi provem entw ith categorical observations. Th is article sets up amathem aticalmodel ofw afer lapp ing and presents a novel algorithm by introducing categorical observations to feedforw ard control system. In add ition, w e verify the capacity of th is algorithm by smi u lation study, usingM SE as the criterion. R esults show an outstand ing performance under the cond ition w ithout orw ith m isclassification ofw afer thickness. Key wo rds: statistical process contro;l feedforw ard contro;l m in mi um mean sum of square; categorical data; sem iconductorm anufacturing 0 , , , , : ; [ 1] , , ; , , , , , , [ 2] , , , : : 国家自然科学基金资助项目( 7080203 ) : 韩凯, 男, 硕士研究生; 王凯 , 男, 副教授, 主要研究方向: 统计质量控制数据驱动的复杂制造系统建模 /监视 /控制与诊断 172 运 筹 与 管 理 2010年第 19卷 , , , , , , , , , , [ 3] , : ( 1) , , ( 2) , ( 3), ( ), ( 5) , [ ] [ 5] , R 2R , , , ( ), 1 , , ,