奇异值小波降噪法在柱塞泵振动信号处理中的应用

针对小波阈值和奇异值分解降噪法的不足,研究一种新的小波阈值函数。提出一种基于改进阈值的奇异值小波降噪方法,该方法利用奇异值分解技术,将噪声非均匀分布的信号正交分解为噪声分布相对均匀的分量,并对每个分量进行小波阈值降噪,重构降噪后的分量,得到降噪信号。仿真实例证明,该方法与小波软、硬阈值及改进阈值法相比,不仅提高信噪比,而且能够更好地消除高斯噪声。利用该方法对柱塞泵不同状态振动信号进行降噪,结果表明,该方法能有效抑制噪声,为柱塞泵振动信号预处理提供一种更为有效的方法。

APPLICATION OF SINGULAR VALUE AND WAVELET DE-NOISING METHOD IN ANALYZING VIBRATION SIGNAL OF PISTON PUMP

In order to eliminate noise signal of piston pump vibration signal,a new wavelet threshold function was used for the shortcomings of wavelet threshold and singular decomposition de-noising method.An improved de-noising method was put forward by combining singular value decomposition and new wavelet threshold function de-noising method.In the new method,the noise signal is decomposed into noise uniformity vectors,and through wavelet threshold de-noising on every vector,finally,reconstruction signal.Emulation experiment results show that the method has better de-noising effect than wavelet soft,hard and new threshold de-noising methods.Different vibration signal of piston pump was de-noised by the method.The result proves it can reduce the noise of the vibration signal of piston pump efficiently.