基于幂函数型双稳随机共振的故障信号检测方法

Fault signal detection method based on power function type bistable stochastic resonance

在实际的故障诊断中,有用信号经常淹没在噪声中,特征信息提取非常困难。为了提取强噪声背景中的微弱信号,将幂函数型单势阱模型与Gaussian Potential模型相结合提出一种新型的双稳随机共振系统,称为幂函数型双稳随机共振系统。首先,以平均信噪比增益为衡量指标,提出一种寻找最优系统参数组合的算法,使微弱信号、噪声及系统产生最佳的共振效果;然后,基于幂函数型双稳随机共振系统对Levy噪声背景下的仿真信号进行检测;最后提出一种基于小波变换和幂函数型双稳随机共振的微弱信号检测方法并应用于轴承故障信号检测中。仿真实验表明,幂函数型双稳随机共振模型在故障信号检测中是有效和可靠的。

In actual fault diagnosis,useful information is often submerged in heavy noise,so the feature information extraction is very difficult. In order to extract weak signal from strong noise background,in this paper through combining the power function type single potential well model and Gaussian Potential model,a novel bistable stochastic resonance system is proposed,which is called power function type bistable stochastic resonance system. Firstly,the average signal-to-noise ratio gain is taken as the measurement index,an algorithm for searching the optimal system parameter combination is proposed,which makes the weak signal,noise and system produce optimal resonant effect. Then,the power function type bistable stochastic resonance system is used to detect the simulated signal under Levy noise background. Finally,a weak signal detection method based on wavelet transform and power function type bistable stochastic resonance system is proposed and applied to bearing fault signal detection. Simulation experiment results demonstrate that the power function type bistable stochastic resonance system is effective and reliable in fault signal detection.