费用约束下含应力优化的CSADT设计

恒定应力加速退化试验(CSADT,Constant Stress Accelerated Degradation Testing)方案中需要确定的变量有应力水平、各应力下样本量、监测间隔和各应力下监测次数.如何对CSADT的应力水平进行优化目前很少有人探讨.针对这一问题提出一种在费用约束下将上述4个变量进行综合优化的方法.首先用随机过程描述CSADT下产品的性能退化过程,然后以产品在使用条件下p分位寿命可靠度的渐近方差最小为目标,以试验费用不超过预算为约束条件,给出CSADT的4个优化变量,并给出应用该优化方法的仿真算例.最后,优化方案对模型参数偏差的敏感性分析表明,在一定偏差范围内优化结果具有良好的稳健性. 

Planning of CSADT with stress optimization under cost constrained

In constant stress accelerated degradation testing (CSADT)scheme, there are four variables: stress levels, interval of performance inspection, sample size and number of inspection at each stress level need to be determined. However, the issue about how to optimize stress levels is seldom discussed. For optimizing the four variables comprehensively a method was proposed. First stochastic process was used to describe a typical CSADT problem. Next, under the constraint that the total experimental cost does not exceed a predetermined budget, the objective is to minimize the asymptotic variance of the estimation of the reliability of the pth quantile of product-s lifetime under use condition. The optimal plan can give the four variables mentioned above. Then a simulation example was presented to illustrate this method. Finally, sensitivity analyses of optimal plans to misspecifications illuminates that the optimal plan is quiet robust for a moderate departure from the assumed value of parameters.