一种新的无失效数据可靠性评估方法研究

为对高可靠性的电子与机电类产品进行可靠性评估，在无失效数据情况下，提出一种将失效信息引进置信限分析中的修正单侧置信限评估法。通过不同截尾时间对应的产品无失效数据，采用Bayes参数估计得到产品失效率，并根据所提出的方法将时间系数进行等效替换，得到产品可靠度与平均故障间隔时间。以某型机器人步进电机的无失效数据为例，将该方法与其他单侧置信限方法的计算结果进行对比分析，结果表明：在置信水平为0.99条件下，产品平均故障间隔时间为37 080 h，小于其他置信限方法的计算结果；当任务时间达到1 000 h时，采用新方法得到的产品可靠度估计值为0.973. 实例分析验证了新方法在无失效数据产品可靠性评估中的有效性，且可防止产品可靠度与平均故障间隔时间估计结果冒进现象的发生。

Research on a New Reliability Assessment Method for Zero-failure Data

In order to analyze the reliability of the high reliability electronic and electromechanical products, a modified unilateral confidence limit analysis method under the condition of zero-failure data is proposed, which introduces the failure information into the confidence limit analysis. The failure rate of the product is estimated by Bayes parameter, and the data of the product with different truncation times are obtained. The time coefficients are equivalently replaced according to the proposed method, and then the reliability of product and the mean time between failures can be known. Based on the zero-failure data of a robot stepping motor, the calculated results of the proposed method is compared with those of the other unilateral confidence limit methods. The results show that, under the confidence level of 0.99, the mean time between failures of the product is 37 080 h, which is less than the calculated result of other confidence limit methods. The reliability of the product is estimated to be 0.973 by using the proposed new method when the mission time reaches 1 000 h. The example analysis proves the effectiveness of the new method in the reliability evaluation of zero-failure data products, and can prevent the product reliability and the mean time between failures estimation from occurrence.