摘要
针对数控机床可靠性研究中数据抽样问题进行了研究。在通过对简单随机抽样、分层随机抽样和二阶抽样的实例计算之后,证明二阶抽样在抽样样本量为简单随机抽样的1/2时,其样本方差远小于简单随机抽样的样本方差;在样本量与分层随机抽样相同的情况下,其样本方差也要远小于分层随机抽样方差,提高了抽样数据的可信度和精度,从而肯定了二阶抽样方法对于数控机床可靠性研究中数据抽样问题的适用性,节约了抽样成本。
The data sampling method of NC machine＇s reliability is studied. Based on the calculations in simple random sampling, stratified random sampling and two-stage sampling, it is shown that when the sampling size of two-stage sampling is 1/2 that of simple random sampling,the mean square error is much smaller than that of simple random sampling. At the same time, the mean square error of two-stage sampling is also much smaller than that of stratified random sampling when their sampling sizes are equal. It is concluded that the two-stage sampling improves the precision and reliability of data sampling. And it can save the cost of sampling.
出处
《电子产品可靠性与环境试验》
2008年第1期30-33,共4页
Electronic Product Reliability and Environmental Testing
关键词
数控机床
可靠性
二阶抽样
NC machine
reliability
two-stage sampling
