基于SSI模型的加速应力试验定量评估方法

Quantitative evaluation of accelerated stress testing based on SSI

为了解决传统应力-强度干涉模型(SSI,Stress-Strength Interference)不能对加速应力试验(AST,Accelerated Stress Testing)进行可靠性指标随时间变化的定量评估问题,针对应力和强度均引入了随机过程进行描述.首先假设了产品强度退化过程和应力变化分别为漂移布朗运动和对称布朗运动.其次,为了解决Basu的布朗运动SSI模型在零时刻可靠度为1的问题,从工程角度,假设产品设计制造过程与试验或使用过程相互独立,提出了产品可靠性是设计生产决定的固有可靠性以及试验使用过程确定的使用可靠性的串联结果.通过将传统应力强度干涉模型结合Basu的布朗运动SSI模型,建立了与时间相关的应力强度干涉模型IMSDⅠ,并给出了相关评估方法.基于将加速应力试验与加速退化试验相结合,改进了已有的AST试验方法,扩大了试验数据量,最后,通过仿真算例验证了本文模型和评估方法的理论正确性和工程可实施性.

In conducting accelerated stress testing （AST） , to quantitatively evaluate reliability with respect to time by use of stress-strength interference （SSI） model, stochastic processes were introduced to describe stress and strength. Firstly, It was assumed that strength degradation and stress variation is drift and symmetric Brownian motion, respectively. Brownian motion SSI established by Basu involved reliability equals 1 when time equals zero. To overcome this problem, from engineering view, reliability of product was regarded as the result, which inherence reliability determined by design and manufacturing is in series with operation reliability decided by experiments and operation. Based on the traditional SSI and Basu＇s model, time-dependent interference model of strength degradation Ⅰ （IMSD Ⅰ ） was derived. With the combination of AST and accelerated degradation testing（ADT） , data collection method in AST was improved. Finally, simulation example verified the correctness and applicability of the methodology proposed.