摘要
根据齿轮箱传动部件故障机理,利用小波变换多分辨率特性和时频局部化特性,提取出故障特征信号;并利用有效的消噪技术,去除噪声干扰。参考专家经验,给出模糊规则及模糊神经网络模型,实现故障推理。并将小波变换和模糊神经网络应用在上海宝钢热轧机的齿轮故障诊断中。
According on the fault mechanism of driving parts of bear box, and using the multi-resolution feature and time-frequency localization feature of wavelet transform, the fault feature signals are extracted. Besides, the noise interference is eliminated by using available noise eliminating method. Consulting expert experience, fuzzy precept and the model of fuzzy neural network are given, and diagnosis inference is fulfilled. Wavelet transform and fuzzy neural network are applied in fault diagnose of gears in hot reduction product of Baoshan iron and steel incorporated company in Shanghai.
出处
《传感技术学报》
EI
CAS
CSCD
北大核心
2006年第3期672-674,745,共4页
Chinese Journal of Sensors and Actuators
关键词
小波变换
消噪
模糊神经网络
齿轮箱故障诊断
wavelet transform
noise elimination
fuzzy neural network
fault diagnose of bear box
