关键词: 滚动轴承, 奇异谱分析, 变分模态提取, 稀疏最大谐波噪声比解卷积, 风力发电机, 故障诊断

Abstract: This article proposed a fault diagnosis method for wind turbine rolling bearings, aiming to solve the problem of difficult feature extraction in strong noise environments. This method combined singular spectrum analysis optimized variational mode extraction and sparse maximum harmonic-to-noise ratio deconvolution techniques. De noising and outlier removal of the original vibration signal, and singular spectrum analysis was used to optimize variational mode extraction and decompose the signal to capture multi-scale fault features in the signal. The sparse maximum harmonic-to-noise ratio deconvolution algorithm was applied to analyze the decomposed signal, extract the fault characteristic frequencies of rolling bearings, and combine them with the fault characteristic frequency library of bearings to achieve accurate diagnosis of fault types. The experimental results demonstrated that this method can successfully extract the characteristic frequency of bearing faults, significantly improve the accuracy and reliability of fault diagnosis, and provide strong technical support for the health monitoring and maintenance of wind turbines.

Key words: rolling bearing, singular spectrum analysis, variational mode extraction, sparse maximum harmonic-tonoise ratio deconvolution, fault diagnosis, wind power generator