Survival in Dynamic Environments
Survival in Dynamic Environments
复制标题
DOI：
10.1214/ss/1177010132
复制
发表时间：
1995-02
影响因子：
5.7
通讯作者：
N. Singpurwalla
中科院分区：
文献类型：
--
作者：
N. Singpurwalla
这篇说明性论文概述了一类相对较新的失效模型，包括单变量和多变量，适用于描述在动态环境中运行的物品的寿命。目前使用的许多故障模型都是在运行环境是静态的前提下开发的：这些模型被证明是这里概述的新模型的特例。这些新模型是通过使用适当的随机过程描述潜在的故障导致机制，如退化和磨损而得出的：这是推动它们发展的潜在数学主题。由于它们的普遍性，新的模型应该提供对故障数据的更好描述和对项目生存能力的评估。此外，它们可能标志着一种新的生命测试实验理念，在这种实验中，人们还可以监测管理测试的环境因素。本综述按用于模型开发的各种策略对模型进行了分类，概述了模型背后的重要假设，并提供了相关参考文献的便捷路线图，这些参考文献散布在应用概率、工程、可靠性和生存分析文献中。我们希望本文能为未来生存模型的开发工作方向定下基调，其中故障物理和运行环境的特性起着关键作用。
This expository paper is an overview of a relatively new class of failure models, both univariate and multivariate, that are suitable for describing the lifelength of items that operate in dynamic environments. Many of the currently used failure models are developed under the premise that the operating environment is static: these models turn out to be special cases of the new models that are overviewed here. These new models are derived by describing the underlying failure-causing mechanisms, such as degradation and wear, using suitable stochastic processes: this is the underlying mathematical theme that drives their development. Because of their generality, the new models should pro- vide improved descriptions of failure data and assessments of item survivability. Furthermore, they may signal a new philosophy of life- testing experiments wherein one also monitors the environmental fac- tors that govern the tests. This overview categorizes the models by the various strategies used for their development, outlines the salient as- sumptions underlying them and provides a convenient road map to the pertinent references, which are scattered among the applied probability, engineering, reliability and survival analysis literatures. It is our hope that this paper sets a tone for the direction of future work in the development of models for survival wherein the physics of failure and the characteristics of the operating environment play a key role.