多态系统可靠性分析广义灰色贝叶斯网络模型

针对多态系统可靠性分析过程中部件和系统的故障状态难以准确界定、故障逻辑关系难以精确测度等问题,基于模糊数学和灰色系统理论提出了一种广义灰色贝叶斯网络模型。该模型首先运用含有区间灰数的模糊子集表征部件所处的故障状态;然后通过运用区间灰数表示系统故障发生的条件概率表征多态系统复杂多样的故障逻辑关系;最后基于贝叶斯网络推理算法和区间灰数运算规则,给出系统处于不同故障状态的概率和部件重要度等可靠性特征信息,为不确定条件下的多态系统可靠性分析提供了完整的解决方案。通过对某民航飞机的可脱离应急跟踪设备的案例研究,表明了所提模型的可行性。

Generalized grey Bayesian network model for reliability analysis ofmulti-state system

Engineering systems and their components usually have multiple failure modes and various performance states. However, it is difficult to distinguish one state from another andto detect their logical relationship. A generalized grey Bayesian network for reliability analysis of multi-state system under uncertainty based on fuzzy sets and the greysystem theory is proposed. In this model, different fault states are represented by fuzzy sets with interval grey numbers. Conditional probabilities of system and components areindicated by interval grey numbers, representing the complex fault logical relationships. Finally, reliability characteristic parameters are calculated with the Bayesian networkinference algorithm and operation rules of grey numbers. The feasibility of the proposed model is demonstrated through a case study about a deployable emergency response devicein a civil aircraft.