摘要
由随机回应模型和连续时间马尔可夫链 ,通过双拉普拉斯变换将 RRM与系统组构成为一种新的马尔可夫概率模型 .通过对预备定理的论证成立了估计故障概率 θ的两个基本条件 和 ,由基本定理分别给出了按基本条件建立的θ1 和θ2 的估算公式 ,考虑一种平稳过程分别讨论了两种极限监理情形 ( v=1 ,v=m)对θ的双侧估计 ,应用基本定理和克拉姆法则求证了双侧估计不等式 .还根据基本条件对特征量 tω( tωk′,tω k″)进行估计 .在最后评述中给出了网络部件系统θ的双侧估计式 .
The RRM and system,from stochastic repayable model(RRM) and continuous time Markov chain(CTMC),were composed into a new Markov probable model(MPM) through both Laplace transformation. And two basic conditions Ⅰ and Ⅱ of estimative probability θ of trouble were found,through demonstration for preparatory theorems. The basic theorms have respectively given that the estimative formulas for θ 1 and θ 2 are built according to the basic conditions.We discussed, by regarding a smooth process as a consequence, two types of limitational supervisory running circumstances (v=1,v=m) for the both sides estimate and proved the inequality of the the both sides estimate with the basic theorems and the Kelamu rule.We also estimated characteristic root t ω(t ωk′,t ωk″) according to the basic conditions.Finally we appraised and put forward the both sides estimative formula of θ for a network components system.
出处
《兰州大学学报（自然科学版）》
CAS
CSCD
北大核心
2003年第5期14-18,共5页
Journal of Lanzhou University(Natural Sciences)
关键词
概率模型
网络部件
双侧估计
极限监理
基本条件
probability model
network assembly
both sides estimate
limitational superintendent and management
basic condition
