基于比例风险模型的可靠性灵敏度分析

摘要

To describe the dynamic influence of environment factors on the reliability, a method for reliability sensitivity analysis based on proportional risk model (PRM) is proposed by means of the varied environment test data. In this method, a PRM is introduced to relate the reliability to environment factors. According to the test data analysis or experience, one of the environment conditions is chosen as the baseline environment covariant and the PRM is transformed. Treating the indicator as Poisson distribution, the log-likelihood function is transformed into be a generalized linear expression of Poisson variable. By using the generalized linear model for Poisson distribution, the maximum likelihood estimations of the model coefficients are obtained. Thus the influences of environment factors on the reliability of the product are measured quantificationally. With the reliability model, the reliability sensitivity is obtained. The instance analysis shows that the method can be used to analyze the dynamic varying character of reliability with environment factors and is straightforward for engineering application.%为了定量描述产品可靠性随环境因素的动态变化,研究了环境因素对可靠性的影响,提出了一种基于比例风险模型的可靠性灵敏度分析方法.首先利用比例风险模型描述产品可靠性水平与环境因素的定量影响关系.结合试验数据选取基准环境协变量,对比例风险模型进行变换,并利用广义线性模型给出了可靠性模型参数的极大似然估计.进而给出了产品关于不同环境因素的可靠性灵敏度,用于度量可靠性对环境因素的灵敏性.结合实例表明该方法综合利用了不同环境条件下的试验数据,提高了可靠性灵敏度分析精度,可用于分析产品可靠性关于环境因素的动态变化特征.