基于贝叶斯网络和关联规则的航电系统故障诊断  

AVIONICS SYSTEM FAULT DIAGNOSIS BASED ON BAYESIAN NETWORK AND ASSOCIATION RULE

作　　者：王凯 李玄玄 Wang Kai;Li Xuanxuan(College of Electronic Information and Automation,Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学电子信息与自动化学院,天津300300

出　　处：《计算机应用与软件》2023年第3期45-51,148,共8页Computer Applications and Software

基　　金：国家自然科学基金委员会-中国民航局民航联合研究基金项目(U1533201);国家自然科学基金青年科学基金项目(61703406);天津市自然科学基金项目(18JCQNJC05000)。

摘　　要：为了提高机载航线维修的效率,提出利用贝叶斯网络结合关联规则来构建航电系统故障诊断模型。有效利用航电系统的故障树转化为以贝叶斯网络为架构的故障诊断模型结构,并将BITE信息作为节点扩展到结构中,采用Leak Noisy-Or模型简化故障诊断模型条件概率的计算,利用关联规则算法挖掘历史维修数据包含的强关联规则,并综合专家经验完成条件概率参数的学习。以VHF系统为例,使用GeNIe2.3软件对故障诊断模型进行仿真,通过对具体案例的推理、诊断分析,实现了机载航线维修的动态化诊断过程。In order to improve the efficiency of airborne line maintenance,a Bayesian network combined with association rule is proposed to establish the avionics system fault diagnosis model.The fault tree of avionics system was effectively transformed into a fault diagnosis model structure based on Bayesian network,and BITE information were extended into the structure as nodes.The Leak Noisy-Or model was used to simplify the calculation of the conditional probability of the fault diagnosis model.The association rule algorithm was used to mine the strong association rules contained in the historical maintenance data,and the conditional probability parameters learning was completed in combination with expert experience.Taking the VHF system as an example,GeNIe2.3 software was used to simulate the fault diagnosis model,the dynamic diagnosis process of airborne line maintenance was realized by reasoning,diagnosis and analysis of specific cases.

关 键 词：航电系统 贝叶斯网络 关联规则 BITE LEAK Noisy-Or 条件概率 故障诊断 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] TP182