摘要
提出一种结合Hilbert变换与威布尔(Weibull)分布的早期故障特征提取新方法。首先对滚动轴承的振动信号进行Hilbert变换,求出其包络,然后将包络拟合为双参数威布尔分布模型,并求出威布尔分布负对数似然函数,将其作为表征滚动轴承运行状态的特征向量。最后,利用欧氏距离判别法进行滚动轴承的故障识别与分类。实验结果表明所提方法的有效性和可行性。
A new method of incipient fault feature extraction using joint Hilbert transform and Weibull distribution is proposed. Firstly,envelope of vibration signal can be obtained by using Hilbert transform. Then the envelope can be fitted as two-parameter Weibull distribution,and Weibull distribution negative log-likelihood function can be obtained and regarded as the feature vector of the bearing＇s running state. Finally,the fault of rolling bearing can be identified and classified by Euclid distance discriminance. The experiment results show the effectiveness and feasibility of the presented method.
出处
《科学技术与工程》
2010年第17期4163-4167,共5页
Science Technology and Engineering
