基于虚拟失效序列样本的系统实时可靠性研究

针对一类不能获得足够数量的试验和现场使用数据的复杂系统,提出Monte Carlo(MC)实时可靠性估计方法.以系统部件失效时间序列作为样本空间,应用MC法仿真产生更多的虚拟样本,建立了虚拟样本序列下系统实时可靠性估计的一般框架,并研究了MC估计的期望和方差与先验信息之间的关系,得到了具有包含关系的信息条件下,基于较少先验信息的估计的方差不超过拥有更多先验信息的估计的方差.考虑所有部件失效时间序列和导致系统失效的部分部件失效时间序列的先验信息,分别导出了系统实时可靠性模型.最后,通过算例分析,验证了本文不同先验信息下可靠性模型的有效性和方法的合理性.

Real-time reliability of system based on virtual failure sequence samples

For a kind of complex systems which the sufficient data cannot be obtained by testing and field operating, Monte Carlo (MC) measures for real-time reliability estimate are proposed. First, the failure time sequence of components is considered as the sample space, and more virtual samples are generated through the MC simulation measure based on it. Then a general framework for real-time reliability estimator is presented, and the relationship of prior information and variance of an MC estimator is studied. Under the information with a relation of inclusion, the variance of an estimator based on less prior information is less than that of an estimator based on more prior information. Second, the failure time sequence of all or part of the components which contribute to the failure of system is considered, and the real-time reliability models are developed respectively. A numerical example is provided to enlighten the idea.