ABSTRACT
In the technology of condition monitoring and fault diagnosis of mechanical equipment, the research on the theory and method of vibration signal processing is the basis of extracting fault feature and effectively developing fault diagnosis. However, in the running process of rolling bearings, signal acquisition is inevitably interfered by a large number of non-monitoring site vibration, thereby resulting in effective information of submerged bearings; this kind of phenomenon can be observed in the early stage of rolling bearing fault behavior (Su, 2012). Rolling bearing is one of the most widely used mechanical parts in all kinds of rotating machinery, and its running state often directly affects the performance of the whole machine (including precision, reliability, and life span). Rolling bearing is the wearing part; according to statistics, 30% of the rotating machinery fault is caused by the rolling bearing; fault monitoring and diagnosis have been the subject of domestic and foreign technical fields (Wang, 2012). When the bearing undergoes local damage or defect, in the operation process, the damaged parts and other parts collide. This impact will produce attenuation impulse, and the impact force pulse frequency range is relatively wide, so as to stimulate the high frequency vibration of bearings. This natural vibration of high frequency is the carrier of the bearing vibration, and its amplitude should be modulated by the pulse excitation force caused by these defects, so that the ultimate bearing vibration waveform performance is the complex amplitude modulation wave (Wang, 2011). Therefore, the fault characteristic signal can be obtained by using the resonance demodulation method.
