恒应力寿命试验截尾模型参数的最佳估计

对恒定应力加速寿命试验建立了定时截尾与定数截尾参数模型.在诂计模型参数时,考虑定时截尾产品失效数的随机性,引入近似因子、修偏乘子和修偏项,建立起参数估计通用式并逐次进行跟随修偏估计.对定数截尾考虑数据信息随机性短缺,通过变量变换分别建立最优线性无偏估计和简单线性无偏估计在指数分布条件下的式子,并按统计分析考虑随机因数的影响引入综合修偏因子,按抽样数目的多少分别采用修偏后的式子.最后应用实际数据对模型参数逐次修偏估计和不同估计方法对比的最佳估计.

The Parametric Best Estimations of Cut Tail''''s Models in Life Testing By the Constont-stress

This paper built the parametdeic models of definite time cut tail and fixed quantity cut tail in accelerated life testing by constont stress.And thought product′s stale number′s randomness,when estimate model′s parameter,then introduce a approximate factor,a revised deviation′s multiplier and a revised deviation′s,item,and thereupon set up a common posture of paramatric estimations,also went on one by one yevised deviation′s estimation to follow.And thought the random shortages of data messages to fixde quantity cut tail,separately established the best linear unbiased estimate′s and the general linear unbiased estimate′s pastures under the condition of the exponential distribution when throgh the variable transform.And according to stasistic analysis also must consider the impact of the random factor so introduced a synthetic factor of revised deviation,then according to more or less of sampled number respectively adopted the two postures after revised deviation.Finally we used an exampl of data,successively made the estimates of revised deviation for model′s parameters and the best estimate contrast different estimative methods.