基于指数分布不同定时截尾数据的可靠度的置信下限

The Lower Confidence Limit of Reliability for Exponential Life Data with Varying Censoring Time

本文基于指数分布不同定时截尾数据,利用鞍点逼近法给出参数估计的概率分布的近似公式,进而给出可靠度的近似置信下限,并通过数值模拟及实例计算说明本文方法的可行性。

For exponential life data with varying censoring time, this paper gives approximation to the distribution of parameter estimation using saddlepoint approximation. Based on the saddlepoint approximation, this paper provides approximative lower confidence limit of reliability. Finally, the numerical simulation shows the proposed methods are feasible.