A Feature Extraction Method of Rolling Bearing Fault Signal Based on the Singular Spectrum Analysis and Linear Autoregressive Model

摘要

A feature extraction method of rolling bearing fault signal based on the singular spectrum analysis (SSA) and linear autoregressive (AR) model is proposed. The SSA is used to achieve the noise reduction, which has three steps: decompose original signal into multiple components, remove the components which have smaller contribution, and reconstruct the signal. Then, the reconstructed signal is modeled by the linear AR model, and the coefficients of the model are extracted as the characteristics of the signal. Finally, the proposed method is verified by using the experimental data of Case Western Reverse Lab.