Stochastic resonance (SR) and self-induced stochastic resonance (SISR) are two kinds of important dynamical phenomena in the nonlinear system. SR occurs at the frequency of the characteristic signal. However, SISR can occur at a frequency that is included in the excitation. In present, there are volumes of literatures focusing on extracting the bearing fault characteristics from the vibration signal by SR method. However, the occurrence of SISR may result in the fault features misjudgment in SR processing. Through experimental verifications, we find that the interference of SISR is illustrated strongly in the fault characteristics identification. More importantly, the transition from SISR to SR corresponds to the evolution process of bearing state from normal to damage. Therefore, this evolutionary process can not only judge the state of bearing, but also describe the severity of bearing failure. The result is verified by processing the signals of bearing fault with different severity in noise background. They are the most important findings in this work.

中文翻译：

---

轴承故障诊断中的随机共振和自激随机共振

随机共振（SR）和自激随机共振（SISR）是非线性系统中两种重要的动力学现象。SR 出现在特征信号的频率上。但是，SISR 可以在包含在激励中的频率上发生。目前已有大量文献着重于通过SR方法从振动信号中提取轴承故障特征。然而，SISR的发生可能会导致SR处理中的故障特征误判。通过实验验证，我们发现SISR的干扰在故障特征识别中得到了强烈的说明。更重要的是，从SISR到SR的转变对应于轴承状态从正常到损坏的演变过程。因此，这个进化过程不仅可以判断轴承的状态，还要描述轴承故障的严重程度。通过在噪声背景下对不同严重程度的轴承故障信号进行处理，对结果进行了验证。它们是这项工作中最重要的发现。