大样本顺序统计量均值方差和协方差计算与验证.docVIP

Ξ大样本顺序统计量均值 、方差和协方差计算与验证COMPUTATIO N ANΞ大样本顺序统计量均值 、方差和协方差计算与验证COMPUTATIO N AND VERIFICATIO N OF MEANS , VARIANCESAND CO VARIANCES FO R L ARGE SAMPL E O RD ER STATISTICS傅惠民ΞΞ林逢春(北京航空航天大学 小样本技术研究中心 ,北京 100083)FU HuiMin LIN Fe ngChun( Research Center of Small S ample Technology , Beijing University of Aeronautics and Astronautics , Beijing 100083 , China)摘要 顺序统计量在可靠性评估和寿命预测中有广泛的应用 。目前 ,样本容量较小时的顺序统计量均值 、方差和协方差的数值计算已得到很好的解决 ,但是 ,由于计算精度问题 ,样本容量较大时的顺序统计量均值 、方差和协方差的数值 计算尚未得到完全解决 。对此 ,文中首先从理论上推导出特殊情况下极值分布顺序统计量均值和方差计算的解析表达式 ,以及正态分布顺序统计量均值 、方差和协方差的简化计算公式 ,为数值计算结果提供验证依据 。然后 ,通过大量数值计算与验证表明 ,目前微机的运算速度与精度已经可以实现大样本顺序统计量均值 、方差和协方差的高精度计算 。文中 还给出部分顺序统计量数字特征的积分计算与验证结果 。关键词 顺序统计量 均值 方差 协方差 可靠性 寿命预测中图分类号 O241. 7 TB115Abstract Order statistics are widely used in reliability evaluation and life prediction. The numerical computation of means , var2i ances and covariances of order statistics has been well solved for small sample , however , it has not yet been completely solved for large sample because of computation precision. Formulas of mean and variance of the first order statistic from extreme value distribution are de2 rived in analytic form. Furthermore , simplified formulas of mean , variance and covariance of normal order statistics are also derived in particular cases. They offer the verification for numerical computation results. A large number of computations and verifications indicate that the speed and the precision of current computers are enough to satisfy the high2precision computation of means , variances and cova2 riances for large sample order statistics. Some computation and verification results are presented.Key words Order statistic ; Means ; Variances ; Covariances ; Relia bility ; Life predictionCorresponding author : FU HuiMin , E2mail : f uhuimin @263. net , Fax : + 86 210 282328501The project supported by the National Natural Science Foundation of China(No . . Manuscript received,in revise