On the use of EEMD and SVM based approach for bearing fault diagnosis of wind turbine gearbox

Abstract

To efficiently identifying the bearing fault in gear-rotor-bearing system of wind turbines, this paper presented an algorithmic solution to carry out the analysis of vibration signals of bearings by the use of ensemble empirical mode decomposition (EEMD) and support vector machine (SVM). Through appropriately decomposing the obtained original vibration signals into a collection of intrinsic mode functions (IMFs) using EEMD, the significant IMFs can be selected in respect to the exacted features based on the energy ratio and approximate entropy. As a result, an early fault diagnosis model can be finally derived by the use of support vector machine (SVM) approach. The proposed algorithmic solution is assessed through simulation experiments and the numerical result validates that the solution can efficiently identify the early faults accurately with sufficient robustness.