摘要
在阐述了小波变换和BP神经网络概念的基础上，根据小波神经网络故障诊断的基本思想，提出了一种基于“能量-故障”的小波预处理神经网络航空发动机诊断方法。实验仿真结果表明，使用该方法提取故障特征加快了神经网络的训练速度，能迅速地进行故障的诊断。
The aeroengine fault diagnosis method was presented, based on the basic idea of wavelet transform and neural network. The proposed method used wavelet transform as the preprocessor of BPNN. The result of simulation showed that the fault could be detected quickly and the training speed of BPNN was dramatically accelerated.
出处
《海军航空工程学院学报》
2009年第3期241-244,共4页
Journal of Naval Aeronautical and Astronautical University
基金
海军航空工程学院基础研究基金(08JCJJ0017)
关键词
小波变换
BP神经网络
故障诊断
wavelet transform
BP neural network
fault diagnosis
