A Fault Diagnosis Approach for Roller Bearings Based on Hilbert-Huang Transform and AR Model
一种基于Hilbert-Huang变换和AR模型的滚动轴承故障诊断方法

A fault diagnosis approach for roller bearings based on Hilbert-Huang transform and AR model is proposed. The Hilbert-Huang transform is used to decompose the vibration signal of a roller bearing into a number of IMF components and the instantaneous amplitudes and frequencies of each IMF component are obtained. Then the AR model of each instantaneous amplitude and frequency sequence is established. The main auto-regressive parameters and the variances of remnant are regarded as the feature vectors. Thus, the Mahalanobis distance criterion function is established to identify the condition and fault pattern of a roller bearing. Practical examples demonstrate that the approach based on Hilbert-Huang transform and AR model can be applied to the roller bearing fault diagnosis effectively.