针对滚动轴承这种非平稳振动信号采用的小波包分解的方法来检测故障的存在，运用神经网络来实现故障的分类，还结合D-S理论融合了多个传感器的诊断结果，提高了故障诊断的准确性并通过实验仿真证实。(This course's job is to use the wavelet packet decomposition method for non-stationary vibration signals of rolling bearings to detect the presence of fault

更改参数可以获得不同大小缺陷的仿真信号，并且画出相应的图像，可以用于轴承故障定量诊断(The simulation signals of different size defects can be obtained by changing the parameters.)