基于噪声辅助信号特征增强的滚动轴承早期故障诊断

针对滚动轴承振动信号特征提取在滤除干扰噪声的同时会将部分有用信号滤除,造成特征信号丢失的问题,提出了一种基于噪声辅助信号特征增强的滚动轴承早期故障诊断方法。采用广义多尺度排列熵筛选准则筛选振动信号,并通过粒子群优化算法优化Duffing振子系统参数,实现Duffing振子系统、输入信号与噪声间的最优匹配,从而提高随机共振效果,将部分背景噪声能量转移到滚动轴承早期微弱故障信号特征上,实现了早期微弱故障信号特征的增强。将所提方法应用于滚动轴承全寿命状态早期故障诊断,并与基于VMD的自适应形态学方法相比较,结果表明了该方法的有效性和可行性。

Early fault diagnosis for rolling bearing based on noise-assisted signal feature enhancement

An early fault diagnosis method for rolling bearing based on the feature enhancement of noise-assisted signal was proposed, aiming at the fault feature of rolling bearing vibration signal, some useful signals are filtered out while filtering out the interference noise, which causes the loss of characteristic signal. The generalized multi-scale permutation entropy screening criterion was used to screen the vibration signal, and the parameters of the Duffing vibration subsystem were optimized by the particle swarm optimization algorithm to achieve the optimal matching among the Duffing vibration subsystem and the input signal and noise, thereby improving the stochastic resonance effect. Part of the background noise energy is transferred to the early weak fault signal feature of the rolling bearing, which enhances the feature of the early weak fault signal. The proposed method was applied to the early fault diagnosis of rolling bearing life state, and compared with the adaptive morphology method based on variational mode decomposition. The results show the effectiveness and feasibility of the proposed method.