基于非均匀周期采样的随机共振研究

Stochastic resonance based on periodic non-uniform sampling

随机共振的参数调节过程比较复杂,而且一般要求采样频率不低于信号频率的50倍以上才能实现随机共振。借鉴非均匀采样的思想,提出基于非均匀周期采样的随机共振实现方法,对不同采样频率下的采样数据进行随机共振,并利用非均匀采样傅里叶变换进行频谱叠加分析,可以实现只需简单确定随机共振参数即可得到特征信号成分。该方法还可以在采样频率小于50倍信号频率时,依然能够得到特征信号频率成分。数值仿真和故障模拟实验研究验证了该方法有效的应用性。

Adjustment process of parameters for a stochastic resonance （SR） is quite complicated. And to achieve stochastic resonance, the sampling frequency is more than 50 times of the signal frequency . By means of the idea of non- uniform sampling, a method of periodic non-uniform sampling was proposed here. The sampling data in different sampling frequencies were fi_rstly operated through SR, and then were processed with non-uniform sampling Flourier.transformation. The proposed method could give the feature signal components in spectral analysis by simply determining a group of resonance parameters. It could also provide the feature signal frequency components when the sampling frequency was less than 50 times of the signal frequency. The numerical simulation and the fault diagnosis tests verified the potential applicability of this method.