基于Signature和Ⅱ型双删失系统寿命数据的统计推断

Statistical Inference Based on Lifetime Data of Signature and Doubly Type-Ⅱ Censored Systems

基于Signature理论和方法,建立Ⅱ型双删失情形下元件寿命的统计推断模型,用极大似然估计及其渐近理论,推导出系统元件寿命分布参数的点估计及渐近置信区间估计的解析表达式,并通过3种优化算法获得相关估计的数值解.仿真结果表明,基于Signature的似然函数统计推断模型对Ⅱ型双删失情形下元件寿命的推断有效.

We established a statistical inference model of the component lif etime for the doubly type-Ⅱcensored cases based on the theory and methods of Signature.By using the maximum likelihood estimation and its asymptotic theory,we derived the analytical expression for the point estimation and asymptotic confidence interval estimation of the component lifetime distribut ion parameters and obtained the numerical solutions of the relevant estimation by three optimization algorithms.Simulation results show that the likelihood function statistical inference model based on Signature is effective for inference of the comp onent lifetime under the doubly typeⅡ-censored cases.