基于 EMD-Hankel-SVD 的高速列车万向轴动不平衡检测

摘要

针对EMD（Empirical Model Decomposition）存在模式频率混叠带来的频谱杂乱的根本缺陷，提出一种高速列车万向轴动不平衡动态检测的新方法。该方法的核心是：对万向节安装机座的振动信号进行 EMD 分解得到基本模式分量，应用基本模式分量信号来构造 Hankel 矩阵，对该矩阵进行奇异值正交化分解，以奇异值关键叠层作为奇异值的选择准则对信号进行重构，应用重构信号的傅里叶谱来检测高速列车万向轴的动不平衡，消除 EMD 分解模式频率混叠带来频谱杂乱，提高了谱的清晰度，凸显了故障特征。应用万向轴动不平衡试验数据对该方法进行试验验证，结果表明：该方法能够有效检测万向轴动不平衡引起故障特征和万向轴的固有振动特性，与纯 EMD 方法相比，该方法在谱的清晰度和故障表征力上得到了显著提高。%A new method of detecting dynamic imbalance due to cardan shaft in high-speed train was proposed applying a synthesized method of EMD (Empirical Modal Decomposition),Hankel matrix and SVD (singular value decomposition)against the aliasing defect between adjacent intrinsic model functions existing in EMD.The vibration signals of gimbal installed base were decomposed through EMD to get the different intrinsic model components.The Hankel matrix,constructed of the single decomposition model components,was orthogonally decomposed through SVD and the key singular values were selected to reconstruct the vibration signal on the base of the key stack of singular values.The Fourier spectrum of the reconstructed signal was applied to detect dynamic imbalance of shaft and the clutter spectrum caused by the aliasing defect between adjacent intrinsic model functions was eliminated so as to highlight the failure characteristics.The method was verified by test data in the condition of dynamic imbalance.The results show the method can effectively detect the fault vibration characteristics caused by cardan shaft dynamic imbalance and extract the natural vibration features.By comparison to the simple EMD,the clarity and failure characterization ability are significantly improved.