摘要
结合HHT与GA-BP神经网络的优点,提出将二者结合用于轴承故障诊断的新方法,并且应用最近提出的方法改进HHT,使其应用更为有效。利用HHT构造出代表振动信号特征的"能-频分布";根据GA-BP网络能够逼近任意非线性函数和具有高效寻找全局最优的特点作为特征分类器,进行轴承故障诊断。该方法应用时不需要对信号进行预处理,也不需要精确计算滚动轴承的故障特征频率。实验结果表明,该方法是可行有效的。研究结果为滚动轴承和其他机械设备的故障诊断提供了新的思路。
Based on the advantages of Hilbert-Huang transform（HHT） and GA-BP neural network（GA-BPNN）,a new method of fault diagnosis for rolling element bearings was put forward by combining HHT and GA-BPNN.The methods for the improvement of HHT put forward recently were applied to make HHT more efficient.The energy-frequency distributions of vibration signals were built by HHT,served as feature vectors of rolling element bearings vibration signals.GA-BPNN acted as a classifier for fault diagnosis based on the fact that it could well approach any nonlinear continuous function and had high efficiency in finding the optimization approach reflecting dynamic features of the systems.This method neither need vibration signals data preprocessing nor the calculation of the precise fault characteristic frequency of the bearing.The experimental results indicate that this method is efficient and valid.
出处
《机床与液压》
北大核心
2010年第17期133-137,共5页
Machine Tool & Hydraulics
关键词
滚动轴承
HHT
GA-BP神经网络
故障诊断
Rolling element bearings
Hilbert-Huang transform
GA-BP neural networks
Fault diagnosis
