基于FTA与BP神经网络结合的某型飞机冷气系统故障诊断  

Diagnosis of Air Conditioning System of an Aircraft Based on FTA and BP Neural Network

作　　者：李锋 陈振 李晨旭 刘麦良 LI Feng;CHEN Zhen;LI Chen-xu;LIU Mai-liang(School of Aircraft,Xi an Aviation University,Xi an 710089,China)

机构地区：[1]西安航空学院飞行器学院,陕西西安710089

出　　处：《液压气动与密封》2023年第11期108-111,共4页Hydraulics Pneumatics & Seals

基　　金：陕西省自然科学基础研究计划(2022JQ-503)。

摘　　要：针对某型飞机冷气系统常见的渗漏故障,结合其结构原理图,首先运用FTA(Fault Tree Analysis)方法,分析得到了该型飞机冷气系统渗漏的所有故障模式及最小割集,建立了故障树的结构函数。再根据故障模式的主要类别,筛选、整理出BP(Back Propagation)神经网络的训练样本,运用BP神经网络的方法建立了该型飞机冷气系统渗漏故障诊断模型,并对模型进行了验证。采用“FTA+BP神经网络”相结合的故障诊断方法,克服了FTA与BP神经网络方法单独诊断时的固有缺陷,提高了故障诊断的准确性和高效性,探索了新的故障诊断方法。This paper mainly studied the common leakage fault of the air-conditioning system of an aircraft,combined with its structural principle diagram.Firstly,by using the method of Fault Tree Analysis,all fault modes and minimum cut sets of the leakage of the aircraft s air-conditioning system were analyzed,and the structure function of fault tree was established.Then,according to the main types of failure modes,the training samples of BP(back propagation)neural network were screened and sorted out.The fault diagnosis model of the leakage of the aircraft s air-conditioning system was established by the BP neural network method,and the model was verified.The fault diagnosis method combining FTA and BP neural network used in this paper overcomed the inherent defects of FTA or BP neural network methods,improved the accuracy and efficiency of fault diagnosis,and explored a new fault diagnosis method.

关 键 词：冷气系统 FTA BP神经网络 故障诊断 

分 类 号：TH137[机械工程—机械制造及自动化] TB42[一般工业技术] TP391[自动化与计算机技术—计算机应用技术]