基于改进小波包奇异值法的齿轮泵振动信号去噪  

De-noising of Gear Pump Vibration Signal Based on Improved Wavelet Packet and Singular Value Decomposition

作　　者：何庆飞[1] 姚春江[1] 陈桂明[1] 陈小虎[1] 杨庆[1] 

机构地区：[1]第二炮兵工程学院装备管理工程系,西安710025

出　　处：《机械科学与技术》2012年第9期1445-1448,共4页Mechanical Science and Technology for Aerospace Engineering

基　　金：国防预研基金项目(9140A27020309JB4701);第二炮兵工程学院科技创新基金项目(XY2010JJB38)资助

摘　　要：针对传统基于信号和噪声频谱不同而实现去噪方法的缺陷,研究了小波阈值求解方法和奇异值分解特征选取方法,提出了改进阈值的小波包奇异值去噪方法。该方法将输入信号小波包分解后,进行新阈值去噪,再利用奇异值分解法对重构后的信号进行去噪,改进阈值的小波包奇异值法去噪效果更好。用该方法对齿轮泵的正常状态和侧板磨损故障状态的振动信号进行去噪处理,能有效地抑制噪声。After a new computing method of threshold and a new choosing method ot characteristic value oi singular value were developed, and an improved de-noising method was presented by combining wavelet packet and singular value de-noising method, aiming at the shortcomings of conventional de-noising methods that based on the frequen- cy spectrum difference between noise and signal. In the new method, vibration signals are first decomposed by the wavelet packet, and then the new threshold de-noising on the wavelet packet coefficient is used. Finally, the recon- struction signal is de-noised by the singular value decomposition. Experimental results show that the new method has better de-noising effect than other conventional methods and efficiently reduces the noise of the vibration signal of gear pump, and it provides an effective method for pretreating vibration signal of gear pump.

关 键 词：齿轮泵 振动信号 小波包分解法 奇异值分解法 去噪 

分 类 号：TH137.5[机械工程—机械制造及自动化] TP274[自动化与计算机技术—检测技术与自动化装置]