The vibration signal is one of important bases for fault diagnosis of wind turbine gearbox. In this article, the correlation of vibration signals acquired from different positions of gearbox is analyzed, and the cross-correlation signal is obtained simultaneously. Then, the model of adaptive filtering by utilizing the cross-correlation signal as a reference signal is established, and the similar components of different vibration signals are extracted. Afterwards, the amplitude spectrum method is utilized to analyze the similar components, and the rotating frequency and its harmonics, which are significant for fault diagnosis of the gearbox, are obtained. Finally, the effectiveness of the method is verified by online vibration data.%振动信号作为风电机组齿轮箱故障诊断的重要依据，一直是风电机组故障诊断领域关注的研究热点。基于不同频率、不同能量振动信号的传播规律，对齿轮箱上不同测点的振动信号的相关性进行分析，建立以互相关信号为参考信号的自适应滤波模型，利用该模型提取振动信号中的相似成分，进而利用幅值谱分析相似成分，得到更加清晰的转频及其倍频，为齿轮箱的故障诊断提供依据。方法的有效性得到实测振动数据的验证。
展开▼
