Event-Altered Rate Models for General Reliability Analysis
Event-Altered Rate Models for General Reliability Analysis
复制标题
用于一般可靠性分析的事件改变率模型
DOI：
10.1109/tr.1979.5220648
复制
发表时间：
1979
影响因子：
5.9
通讯作者：
P. Moranda
中科院分区：
文献类型：
--
作者：
P. Moranda
一般泊松家族中的随机过程，但其特征在于具有随事件而变化的速率，而不是随时间而变化，准确地描述了软件错误检测过程。作者已经开发了几种有前途的新模型，并将其应用于硬件和软件的可靠性问题，特别是在老化和磨损阶段。描述磨损的增长模型适用于软件，因为普遍的补丁，系统不可避免地退化。在事件发生时（例如在软件调试的情况下的错误检测和移除）降低固定量或降低到最近速率的一小部分的故障率被审查，并且描述或建议新的应用。其他相关的模型中，利率增加的描述和一些可能的应用提出了建议。
Stochastic processes within the general Poisson family, but distinguished by having rates which are changed by events, rather than by time, accurately describe the software error detection process. Several promising new models of this general type have been developed by the author and apply to hardware and software reliability problems, especially during the bum-in and wear-out phases. The growth models which describe wear-out apply to software when, because of pervasive patching, the system inexorably degrades. Failure rates which are decreased at the occurrence of an event (such as an error detection and removal in the case of software debugging) by either a fixed amount, or to a fraction of the most recent rate are reviewed, and new applications are described or suggested. Additional related models in which the rates increase are described and some possible applications are suggested.