论文部分内容阅读
在对高可靠性指标产品进行可靠性评估时,为满足试验样本量的需求,对不同阶段或不同批次的试验样本进行一致性检验,已成为扩大试验样本量的一种常规手段。针对成败型产品的试验样本,提出了利用相关分析进行一致性判别的方法,并通过实例将常用的Fisher法与相关分析法进行了深入的对比和分析。通过算例分析表明:相关分析法克服了Fisher检验法精度较差、灵敏度较低等缺点,对成败型产品的试验样本进行一致性检验差别结果更加合理可行。
In the reliability assessment of high reliability index products, in order to meet the demand of test samples, the consistency test of test samples in different stages or different batches has become a conventional measure to expand the sample size. Aiming at the test samples of success or failure products, a method of consistency identification by correlation analysis is proposed, and the common Fisher method and correlation analysis method are compared and analyzed through examples. The case study shows that the correlation analysis method overcomes the shortcomings of Fisher’s test, such as poor accuracy and low sensitivity. The consistency test of the test samples of the success or failure products is more reasonable and feasible.