考虑随机失效阈值的退化数据建模分析方法

Degraded data modeling and analysis method taking account of random failure threshold

针对退化过程中随机失效阈值分布难以确定,无法直接引入到退化数据建模分析过程的难题,提出一种将失效阈值的随机性转移到易于获得分布参数变量中的建模分析方法,给出退化试验数据和加速退化试验数据的建模分析步骤.结合实例分析表明:将失效阈值随机性转移到易于获得分布参数变量中的建模分析方法可行,避免了复杂的数值计算;同时,该方法能够为高可靠长寿命产品提供更符合实际的可靠性评估结果.

Aimed at the problem that the random threshold distribution in degradation process is difficult to be determined and introduced into degraded data modeling and analysis process a modeling and analysis method is presented,where the randomness of failure threshold is transferred into such a variable that its distribution parameter can be easily obtained,and the modeling and analysis procedure of degradation test data and accelerated degradation test data are given out.It is shown by analyzing actual case that foregoing modeling and analysis method is feasible and complex numerical computation will be avoided.Meantime,this method can provide a reality-conforming result of reliability estimation for the products with high reliability and long service life.