基于VEITD和OSMHD的风电机组轴承损伤识别

针对风力发电机轴承损伤信号易被强噪声干扰导致损伤特征提取困难的问题，提出了一种基于可变熵加权融合的固有时间尺度分解和优化稀疏最大谐波噪声比解卷积法相结合的风力发电机轴承损伤识别方法。采用固有时间尺度分解方法对原始信号进行分解，得到若干个固有旋转分量。利用可变熵对固有旋转分量进行加权融合。使用优化稀疏最大谐波噪声比解卷积法对加权融合信号进行处理，提取轴承损伤特征频率。试验台数据和风力发电机现场数据分析结果表明，所提方法对轴承损伤信号中的噪声抑制效果明显，能够准确提取风力发电机轴承损伤特征频率，实现风力发电机轴承的损伤识别。

Wind turbine bearing damage identification based on VEITD and OSMHD

The damage signals of wind turbine bearings are easily to be interfered by strong noise, which makes it difficult to extract damage features. Aiming at this problem, a wind turbine bearing damage identification method is proposed, which combines variable entropy intrinsic time-scale decomposition(VEITD) and optimization sparse maximum harmonics-noise-ratio deconvolution(OSMHD) methods. The original signal is decomposed by the intrinsic time scale decomposition method, by which several intrinsic rotation components are obtained. The intrinsic rotation components are weighted and fused by variable entropy. The weighted and fused signals are processed by the OSMHD method, by which the characteristic frequencies of bearing damage are extracted. The analysis results of test-bed data and field data of wind turbine show that the proposed method has obvious noise suppression effect on the bearing damage signal, and can accurately extract the characteristic frequency of wind turbine bearing damage, so as to realize the damage identification of wind turbine bearings.