中文摘要：

针对大型回转支承工况恶劣、背景噪声高,且振动信号非平稳的特点,提出了一种基于聚类经验模态分解-主成分分析（EEMD-PCA）的降噪方法.通过EEMD和PCA将回转支承整个寿命周期的振动信号与回转支承使用初期的振动信号进行对比,确定多个回转支承振动信号中影响较大的经验模态函数（IMF）,最后进行信号重构,完成降噪过程.为验证降噪效果,利用PCA对降噪信号建立了回转支承性能衰退指标.结果表明,提出的方法比现有方法得到的衰退趋势更接近回转支承实际的衰退过程,为后续寿命预测等研究提供了有效的信号处理方法.

英文摘要：

An ensemble empirical mode decomposition-principle component analysis （EEMD-PCA） nethod was proposed to denoise non-stationary vibration signals with strong white noise generated by large-size slewing bearings.Vibration signals of the whole service life were compared with that of incipient periods using EEMD-PCA,and several significant intrinsic mode functions （IMFs） were selected to reconstruct signals for finishing the denoising process.To verify the proposed method,experiments were conducted and the life cycle vibration signals were denoised.The performance degradation model was established by PCA to explain the denoising effect.Results showed that the proposed method acquired a better denoising effect and provides a potential signal processing method for further research.