采用量子理论的滚动轴承故障诊断方法

Fault Diagnosis Method for Rolling Element Bearing by Using Quantum Theory

为克服传统降噪算法在降噪的同时也降低了有用信息的缺点,借助量子理论特有的叠加态表达方式,提出了融合量子叠加态和概率分布的滚动轴承故障诊断方法。该方法利用Laplace数理统计模型描述振动信号双树复小波系数的概率分布规律,在充分考虑了相邻尺度噪声和故障信息变化特点的基础上,采用量子叠加态对振动信号中的噪声和故障信息进行区分。利用传动系统的滚动轴承故障振动信号对本方法进行验证,结果表明,所提方法在降低噪声的同时也有效增强了故障信息,可准确提取出强噪声背景下的机械故障信息。

In order to overcome the disadvantages of traditional denosing method always reduces useful information while it removes noise,a fault diagnosis method combining quantum superposition and probability statistic applied to rolling element bearing is proposed,which uses the specific expression in quantum theory.This method utilizes Laplace mathematics statistics models to describe the dual-tree complex wavelet transform coefficient distribution of vibration signal.On the basis that the change characteristic of noise and fault information between adjacent scales is studied,then quantum superposition is employed to distinguish fault information from noise in vibration signal.The rolling element bearing signal from transmission system with fault is used to verify the validity of this proposed approach.The results reveal that this method can effectively enhance fault information while it reduces noise,thus it can extract mechanical fault information from intensive background noise.