基于RBF神经网络的A320引气系统故障诊断研究

Research of A320 aircraft bleed system fault diagnosis based on RBF neural network

根据神经网络的非线性、自适应性及自学习等优点,针对空客A320飞机引气系统,提出了一种基于RBF网络的A320引气系统智能故障诊断方法。对近几年记录的A320引气系统故障表现与故障原因进行特征提取和分类,并用于RBF网络的学习。利用部分典型故障特征对网络进行测试表明,该故障诊断方法可以有效地根据空客A320飞机的引气系统常见故障表现判别出故障的原因。

According to the advantage such as nonlinearity,adaptive and self-learning of neural networks,aiming at the bleed system of airbus A320,this thesis proposes an intelligent fault diagnose approach.By abstract and classify of the fault behavior and fault source of airbus A320 bleed system collected in the past few years,obtain data is used in the training of a RBF network.And test the network with some typical fault indicate that the fault diagnose method could derive the fault source with the fault behavior of airbus A320 bleed system.