基于模糊故障树的计算机硬件故障诊断

Research on Fault Management Based on Fuzzy Fault Tree

随着计算机技术应用的不断深入,需要提出有效诊断计算机硬件故障并对故障类型进行合理分析的方法。文中通过以Linux操作系统环境为研究对象,针对计算机系统硬件故障发生概率具有模糊性和不确定性的特点,将模糊理论引入计算机硬件故障分析方法中来,通过引入模糊集的概念将常规故障树中基本事件的发生概率模糊化。在实际分析过程中以CPU错误为例,采用上行法求解最小割集并采用转化三角模糊数的方法,为系统的可靠性评估、故障诊断以及维修决策提供了理论依据和实现方法。

Along with the development of computers technology, it needs to come up with effective fault diagnosis and reasonable analysis method. It takes Linux operating system as research object,due to the system failure probability of components is fuzzy and uncertain, fuzzy numbers used in bTA （ Fault Tree Analysis） method is suitable to analyse the reliability of the Linux computer system. In particular CPU error fault tree, use the up method for solving the minimum cut set, and triangular fuzzy number is helped to scribe the probability of interval representation and calculation methods. The results help to provide academic basis of system reliability evaluation, fault diagnosis and repair decision-making.