PFPN在民航飞机故障诊断中的应用  

Application of PFPN in Fault Diagnosis of Civil Aircraft

作　　者：张鹏[1] 王婷婷[1] 俞利明[1] 

机构地区：[1]中国民航大学工程技术训练中心,天津300300

出　　处：《控制工程》2011年第3期483-486,共4页Control Engineering of China

基　　金：机场航班进离港过程建模和调度理论研究();国家自然科学基金(60472124)

摘　　要：针对民航飞机系统故障诊断的复杂性以及民航对快捷性和安全性的要求,采用了基于优先权值的模糊Petri网(PFPN)的故障诊断模型对某型号飞机空调系统进行故障诊断,以提高诊断过程的快速性和准确性。利用概率统计分析数据得到模糊产生式规则并建立模糊petri网,根据正反向推理相结合的算法,在反向推理中引入表示故障严重程度的优先权值,查找故障路径,用正向推理计算故障发生的可信度,得到故障起始原因,并检验反向推理得到的路径有效性。实验结果表明,增加的优先权值参数解决了当同一级别的几个变迁的可信度相差不大甚至相等时,仅依据可信度的最大的原则进行取舍的不合理性的问题,并且反正向相结合的推理算法提高了推理搜索速度,验证了算法的有效性。For the complexity of civil aircraft system fault diagnosis and Civil＇ s requirements of fast and security, this paper proposes a model of fault diagnosis based on fuzzy Petri nets introducing priority values to improve quickness and accuracy of diagnostic process. The datas are analyzed by probability and statistics to generate fuzzy production rules and to establish fuzzy Petri nets. Based on the backward-forward reasoning algorithm, In backward reasoning, the priority value representing the severity degree of fault is introduced to find the fault path, and the forward reasoning algorithm is used to calculate the real value of the fault, obtain the initial cause of fault, and test the reliability of the path of the fault propagation. The results show that the priority value can solve the question that when the real values are approximately equivalent or even equivalent, only taking the values of certainty factors into account is obviously unreasonable, and the backward-forward reasoning algorithm increases the speed of reasoning research, so that it verifies effectiveness of the algorithm.

关 键 词：飞机系统 故障诊断 模糊PETRI网 优先权值 推理算法 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]