指数分布冷贮备系统产品的统计分析——转换开关指数型的情形

The Statistical Analysis of the Exponential Distribution for Cold Standby System——Transfer Switch with Exponential Distribution

给出了全样本场合下指数分布冷贮备系统产品寿命分布中参数θ≠λ时的矩估计和极大似然估计,通过Monte-Carlo给出了参数矩估计的精度,考察了1000次满足条件时所需要的模拟次数,随着样本量的增大,矩估计存在的比率逐渐增大,而极大似然估计的结果与样本有关.同时给出了参数θ=λ时的矩估计、极大似然估计和逆矩估计,通过Monte-Carlo模拟考察了参数点估计精度,认为矩估计比较优.文章还给出了求参数区间估计的两种方法——精确方法和近似方法,通过Monte-Carlo模拟认为精确方法精度较高.

Moment estimate,maximum likelihood estimate are derived in the exponential distribution for cold standby system with full sample size when the parameter θ is not equal to the parameter λ.The moment estimate and the total simulations number are obtained by using Monte-Carlo simulations when the conditions are satisfied.With the increasing of the sample size,the existence percentage of the moment estimations increases.And the result of the maximum likelihood estimate is related to the samples＇ values.In addition,moment estimate,maximum likelihood estimate and inverse moment estimate are derived when the parameter θ is equal to parameter λ.The moment estimation is verified to be superior to other estimations by using Monte-Carlo simulations.Thereafter we use two methods-the accurate estimation and the approximate estimation to obtain the interval estimations of the parameters.The accurate estimation is proved to be better through Monte-Carlo simulations.