Formal Support for Quantitative Analysis of Residual Risks in Safety-Critical Systems
Formal Support for Quantitative Analysis of Residual Risks in Safety-Critical Systems
复制标题
正式支持安全关键系统中残留风险的定量分析
DOI：
--
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
S. Nadjm
中科院分区：
文献类型：
--
作者：
Jonas Elmqvist;S. Nadjm
随着安全关键系统中软件和电子设备的复杂性不断增加，在保持高度保证的同时降低成本和缩短上市时间的新挑战已经出现。在安全评估过程中，目标是最小化风险，特别是可能的故障对系统级安全的影响。必须识别和分析每个潜在故障，以确定最需要关注的故障。在本文中，我们通过容错的定量分析扩展了我们早期的正式定性分析工作。我们的分析基于正在构建的系统的设计模型。它进一步建立在故障的正式模型的基础上，该模型已针对估计的发生概率进行了扩展，从而允许分析系统级故障概率。这是在概率模型检查器 PRISM 的帮助下完成的。该扩展改进了成本高昂的认证过程，其中必须评估所有可预见的故障对安全性和可靠性的影响。我们使用航空电子行业的应用程序演示我们的方法：高度计系统。
With the increasing complexity in software and electronics in safety-critical systems new challenges to lower the costs and decrease time-to-market, while preserving high assurance have emerged. During the safety assessment process, the goal is to minimize the risk and particular, the impact of probable faults on system level safety. Every potential fault must be identified and analysed in order to determine which faults that are most important to focus on. In this paper, we extend our earlier work on formal qualitative analysis with a quantitative analysis of fault tolerance. Our analysis is based on design models of the system under construction. It further builds on formal models of faults that have been extended for estimated occurence probability allowing to analyse the system-level failure probability. This is done with the help of the probabilistic model checker PRISM. The extension provides an improvement in the costly process of certification in which all forseen faults have to be evaluated with respect to their impact on safety and reliability. We demonstrate our approach using an application from the avionic industry: an Altitude Meter System.