基于变分模态分解的轴承振动特征提取方法

Rolling Bearing Vibration Feature Extraction Based on VMD

设备运转的状态信息能够通过振动信号实时反映出来,然而由于信号中混杂了大量背景噪声等干扰信息,使得信号分解技术成为关注的重点之一。变分模态分解(variational mode decomposition,VMD)克服了传统自适应信号分解方法的不足,分解出的信号消除了端点效应和模态混叠等失真现象,具有抗噪干扰能力强、计算速度快等优点。针对VMD模态K数难以选取的问题,以信号主频率个数作为K的选择依据,然后结合信息熵测度,提出了一种的新的振动信号提取方法,剔除干扰信息,便于故障类型的查找。仿真和轴承实验表明了该方法的有效性和可行性。

The information of equipment operation can be reflected in real time by the vibration signal.However,the signal decomposition technology is one of the focuses because the interference information such as a lot of background noise can be in the signal.The variational mode decomposition(VMD)overcomes the shortcomings of the traditional adaptive signal decomposition method.The decomposed signal eliminates the distortion of the endpoint effect and modal aliasing,which has the ability of anti-noise interference ability,the fast calculation speed and so on.In order to solve the problem that the number of K numbers in VMD mode was difficult to be selected,the number of main frequency of signal was chosen as K.Then,a new vibration signal extraction method was proposed to eliminate the interference information and find the fault type.The simulation and bearing experiments showed the effectiveness and feasibility of the proposed method.