神经网络的滚动轴承故障诊断

本资料包含pdf文件1个，下载需要1积分

Through the rotating machinery fault experimental platform, the vibration signals of the rotating machinery fault experimental platform under three conditions such as bearing normal, bearing inner ring cracks and bearing outer ring cracks were collected. After the signal was zero - mean processed ; the signals with the main frequency band of the vibration signals were restructured; the dimensionless time domain as characteristic value was extracted; the fault types by neural networks were recognized. A good fault diagnosis result was obtained in experiments.文章通过旋转机械故障实验平台，采集旋转机械故障实验台轴承的3种工作状态分别是轴承正常、轴承内圈裂缝、轴承外圈裂缝的振动加速度信号．对信号进行零均值化处理后，选择频率成分幅值较大的频率进行信号重组，提取其时域量纲特征值，利用神经网络进行故障类型的识别；通过实验，取得了很好的诊断结果．