论文部分内容阅读
对滚动轴承振动信号进行小波包分解,提取频带能量特征构成特征向量,并以此作为BP神经网络的输入,对神经网络进行训练,建立滚动轴承运行状态分类器,用以识别滚动轴承的运行状态。试验结果表明,通过小波包分解提取能量特征结合BP神经网络对滚动轴承进行故障诊断的方法是可靠的,可以准确识别轴承的故障类别。
The vibration signal of rolling bearing is decomposed by wavelet packet, the energy characteristic of the frequency band is extracted to form the eigenvector, and then the BP neural network is used as input to train the neural network. The rolling bearing operating state classifier is established to identify the running state of the rolling bearing. The experimental results show that the method of fault diagnosis of rolling bearings through wavelet packet decomposition and energy feature extraction combined with BP neural network is reliable and can accurately identify the bearing fault category.