融合非线性加速退化模型与失效率模型的产品寿命预测方法

针对高可靠性产品寿命数据少、获取成本高的问题，基于充分利用产品在研制、加速试验等不同环境下的退化数据、失效数据等可靠性数据的思想，提出了一种融合非线性加速退化模型和失效率模型的产品寿命预测方法.首先，根据退化数据对非线性退化过程进行分析，估计退化过程的参数；然后，根据加速退化数据及相应的加速退化模型估计加速退化模型的参数，从而得到退化参数与应力之间的关系.进一步，利用比例风险模型融合产品的寿命数据和未失效截尾数据，并基于此计算产品的可靠度函数、预测产品的寿命.实例应用验证了所提方法的有效性，同时说明了所提方法的应用价值.

Life Prediction Approach by Integrating Nonlinear Accelerated Degradation Model and Hazard Rate Model

It is difficult to predict the life of highly-reliable products when the failure data are rare and the testing is costly.As such,based on the degradation data and failure data in the developing or accelerate testing stages,a life prediction method is proposed by integrating nonlinear accelerate degradation model and hazard rate model.The degradation process is first modeled as a nonlinear degradation process and the parameter estimation is achieved by analyzing the degradation process.Then an accelerate model is adopted and its parameters are estimated based on the accelerated degradation data to build the relationship between the parameters of the degradation model and stress levels.Further,the Cox proportional hazard model is adopted to fuse the failure data and censored data.To do so,the reliability function can be computed and the life of the product can be predicted accordingly.Finally,the usefulness and practical value of the presented method is demonstrated through the case application.