摘要
对于高可靠长寿命产品,利用产品的退化数据进行可靠性评估的一种常用方法是基于退化量分布的方法。这种方法需要假设产品各检测时刻退化量分布服从相同的分布形式,这在实际中往往不能满足。本文提出的方法充分利用各检测时刻退化数据的信息,利用模型选择的方法获得产品在各检测时刻的退化量分布,在不依赖分布参数的退化轨迹前提下,通过分布拟合估计产品寿命分布参数进行可靠性评估。最后通过实例分析,验证了本方法的灵活性和有效性。
A common approach of reliability evaluation of high-reliability and long-lifetime products using degradation data is the approach based on degradation measure distribution. The assumption that degradation measure distribution at each inspection time follows the same distribution form needs to be made using this approach, which isn＇t hold sometimes in practice. The approach presented in this paper takes full advantage of information of degradation data at each inspection time. the degradation measure distribution at each inspection time is got using model selection, then the values of unknown parameters of life distribution are got through distribution fitting and reliability valuation is performed being independent of degradation measure distribution parameters＇ degradation paths. Finally, this paper gives an example to demonstrates the feasibility and validity of this approach.
出处
《系统工程》
CSCD
北大核心
2009年第11期111-114,共4页
Systems Engineering
关键词
退化数据
模型选择
三参数成布尔分布
最小二乘法估计
Degradation Data
Model Selection
Three-parameter Weibull Distribution
Least Squares Estimate
