基于二维互补随机共振的轴承故障诊断方法研究

一维随机共振(One-Dimensional Stochastic Resonance,1DSR)被广泛用于轴承故障诊断中。针对传统1DSR对微弱信号的检测效果不够理想,输出信号噪声大,不能准确获得轴承故障特征频率(Fault Characteristic Frequency,FCF)等问题,提出一种新的二维互补随机共振(Two-Dimensional Complementary Stochastic Resonance,2DCSR)方法并应用于轴承故障诊断。将采集到的轴承故障信号根据共振带位置进行带通滤波并解调,随后将解调信号对半分成两个子信号并输入2DCSR的两个输入端,利用输出信号的加权功率谱峭度(WPSK)指标对2DCSR系统参数进行自适应调节优化,得到最优的滤波输出信号及频谱,以识别轴承FCF并诊断轴承故障类型。数值仿真及实验结果表明,提出的方法可以有效地增强轴承FCF并提高轴承故障诊断效果。

Bearings fault diagnosis based on two-dimensional complementary stochastic resonance

One-dimensional stochastic resonance (1DSR) methods have been widely used in bearing fault diagnosis. However, some deficiencies still exist in the traditional 1DSR methods such as limited weak signal detection capacity, obvious output noise and inaccurate bearing fault characteristic frequency (FCF) for fault recognition, etc. To address these issues, this study proposes a new two-dimensional complementary stochastic resonance (2DCSR) method to enhance bearing fault diagnosis. First, the acquired bearing fault signal is bandpass-filtered according to the location of resonance band and then demodulated. Then the demodulated signal is split into two sub-signals and the sub-signals are sent to the two input channels of the 2DCSR. The weighted power spectral kurtosis (WPSK) of the output signal is used as the criterion to adaptively guide parameters tuning in the 2DCSR system. Finally the optimal output signal and its spectrum are obtained for bearing fault recognition. Numerical and experimental results show that the bearing FCF can be enhanced by the proposed 2DCSR method, thereby improving the performance of bearing fault diagnosis. Key words: bearing fault diagnosis; two-dimensional complementary stochastic resonance; weighted power spectral kurtosis; weak signal detection