可靠性仿真检验方法

Test method for reliability simulation

产品寿命的不断提高使得可靠性试验费用大、时间长,工程上往往难以承受,同时随着数字化设计、智能制造的快速发展,都对基于数字仿真的可靠性测试需求越来越迫切。但是可靠性仿真的可信性问题长期困扰着人们,为此,本文建立了一种可靠性仿真检验方法,能够以高置信水平判断产品可靠度和可靠寿命的仿真是否正确。同时,鉴于产品性能退化是其可靠性降低并逐步失效的主要原因,文中还给出性能退化曲线可靠性仿真的检验方法,能够以高置信水平判断性能退化曲线在哪一时间段已被正确仿真,在哪一时间段尚未被正确模拟,为进一步分析性能退化曲线未被正确仿真的原因提供指导。本文方法不但对正态分布、Weibull分布、极值分布、指数分布、二项分布等任意分布都适用,而且对完全数据、不完全数据和无失效数据等各种类型的数据也均适用。并且计算简单,便于工程应用。

High cost and long duration of reliability experiment arisen by the improvement of product lifetime,as well as the rapid development of digital designing and intelligent manufacturing,lead to the urgent demand for reliability test based on digital simulation.To settle the credibility issue of simulation,a test method for reliability simulation is proposed in this paper to judge whether the reliability and reliable lifetime of product are simulated correctly with high confidence level.Considering the performance degradation is the main reason for product reliability reduction and even failure,a test method for reliability simulation of degradation curve is further presented,which can distinguish the correct simulated segment and the incorrect one with high confidence level.The proposed methodology is applicable to any distribution,as well as a wide variety of data including the complete failure data,incomplete data and zero-failure data.In addition,the easy calculation also makes it convenient for engineering application.