摘要
为提高飞机航电系统故障诊断效率,充分利用故障模型中的概率信息,减少故障诊断时的计算量,使用概率因果网络中的连接树算法建立故障模型,使故障节点中的信息满足全局一致性,利用逆转弧方法(AR)得到信息传递图,确定信息在节点中的传递方式,采用惰性传播及逆转弧变量消除算法(LAZY-ARVE)进行概率计算,得到最终的诊断结果。最后以飞机通信系统出现故障,接入ACARS控制组件的FMGC信号错误为例进行验证。
To improve fault diagnosis efficiency of the aircraft avionics system by making full use of probability information in the fault model and reducing calculation amount in fault diagnosis, this paper uses the junction tree algorithm in the probabilistic causal network to build the fauh model in such a way that the information at fault nodes may meet the global consistency at the fault node. Arc reversal （ AR） is used to obtain the information transitive graph and determine the information transfer mode at the nodes. Lazy propagation and arc reversal variable elimination algorithm（ LAZY-ARVE ） is used to calculate the probability to obtain final diagnosis result. At the end, a wrong FMGC signal access to ACARS control component in the case of faulty aircraft communication system is taken as an example for verification purpose.
出处
《电气自动化》
2014年第4期16-18,34,共4页
Electrical Automation
关键词
因果网络
连接树
故障诊断
逆转弧
惰性传播
causal network
junction tree
fault diagnosis
arc reversal i lazy propagation
