网络计划中时间均值的方差问题.pdfVIP

第 28 卷 第 6 期 昆 明 理 工 大 学 学 报 ( 理 工 版 ) Vol . 28 　No 6 2003 年 12 月 　　Journal of Kunming University of Science and Technology (Science and Technology) 　Dec . 2003 网络计划中时间均值的方差问题 丁怀 日,谭荣建 ( 昆明理工大学 建筑工程学院 ,云南 昆明　650224) 摘要 : “运筹学”权威教材中关键工序作业时间均值的方差公式推导错误. 一般数理统计中的样 本是直接抽取的简单随机样本 ,即样本所包含的个体之间是相互独立的. 文章所针对的教材都用 直接样本先计算出间接样本 ,再用间接样本计算样本均值及其方差. 这样的间接样本是统计相关 的 ,可是上述教材在计算样本方差和样本均值的方差时 ,又用了随机样本的方差公式 ,所以结果 是错误的. 文中推导的结果竟是原公式的 3 倍. 关键词 : 运筹学 ; 网络计划 ; 间接样本 中图分类号 :O22 文献标识码 :A 文章编号 :1007 - 855X (2003) 06 - 0084 - 03 Variance Questions of Average Val ue Time in Net work Plan DIN G Huairi , TAN Ro n gjian (Faculty of Architectural Engineering , Kunming University of Science and Technology , Kunming 650224 , China) Abstract : There are some errors in variance formula deduction about average time values of critical procedure , in some authoritative “op eration research ”textbooks . Samples in general mathematic and p hysical statistics are simple , which are selected randomly and directly . That means the unit s in these samples are indep endent and have nothing to do with each other . These indirect samples are first calculated based on the direct ones in some textbooks , and then the sample averages and it s variance are calculated based on the indirect samples . These in direct samples have some statistical relationship with each other , but when calculating sample variance and aver age variance , the variance formula of random samples are used , so a wrong result is deduced . The result de duced here is as 3 times as the one deduced by the previous formula . Key words : op eration research ; network pla