Fault detection of rolling bearing based on EMD-DPCA

As one of the most widely used parts and components of rotating machineries, fault detection of rolling bearing is of great significance. In this paper, a new method named EMD-DPCA is proposed based on Empirical Mode Decomposition (EMD) and Dynamic Principal Component Analysis (DPCA). Firstly, the vibration signals are decomposed by EMD and Intrinsic Mode Functions (IMFs) are achieved. Then DPCA model is established by many selected IMFs and used to detect rolling bearing´s failures. The proposed scheme is verified with experimental data collected from deep groove ball bearing of a 2 hp motor driven mechanical system and the results show that the strategies can detect bearing faults efficiently.