Bearing fault diagnosis of wind turbine based on intrinsic time-scale decomposition frequency spectrum

Abstract

In order to better identify the complex running conditions of wind turbine main bearings, we developed a bearing fault diagnosis method based on intrinsic time-scale decomposition frequency spectrum. The main bearing acceleration vibration signal from wind turbine is captured under four conditions—good bearing, outer race fault, inner race fault and roller fault. The proposed method consists of the following steps. First, the main bearing acceleration vibration signal is decomposed into several proper rotation components by using the intrinsic time-scale decomposition method. Second, the frequency spectrum of the first few proper rotation components (containing dominant fault features) is analyzed. The dominant resonant frequency range of each analyzed rotation component is derived, and then, the sum of frequency amplitude in said frequency range can be obtained. This sum is regarded as the fault feature vectors. Finally, the fault feature vectors are input to the least square support vectors machine, and the faults of wind turbine main bearing then can be diagnosed. The experiment results show that the proposed method can diagnose failures of wind turbine bearings quickly and more accurately.