Research On Fault Diagnoses Of Rolling Bearing Based On The Energy Operator Demodulation Approach

Rolling element bearing is widely used in lots of rotating machinery. As one of the key components of a machine, the condition of rolling element bearing has direct influence on the performance of the whole machine. The research on rolling element bearing fault diagnoses technology is very significant. The research area of this dissertation is the detection of defects in rolling element bearing. A system of virtual instruments is designed in order to diagnose all kinds of faults in rolling element bearing.The main research content of this dissertation is shown in the following:(1) The affinity between vibration and noise in rolling element bearing is proved. Considering the shortage of current bearing vibration-measuring instrument in china, this paper puts forward that characteristic parameters to reflect the level of "abnormal noise" must be added. Principium research on abnormal noise is made by concepts and methods of sound signal processing and psychology acoustics.(2) This paper analyzes vibration signals of rolling element bearing in time-domain and frequency-domain, and puts forward the characteristic parameters to reflect the level of vibration and noise.(3) This paper presents the high-frequency resonance technique which is used availability with rolling element bearing fault diagnoses, and introduces a new optimizing arithmetic model, the energy operator optimizing arithmetic; the new arithmetic is compared with the classic arithmetic, Hilbert demodulating method by imitation experiments.(4) Based on LabView language, a system of virtual instruments is designed, including Data acquisition and register instrument, Real-time analysis instrument and Spectrum instrument. At last, this dissertation validates that the high-frequency resonance technique based on energy operator optimizing arithmetic is availability with rolling element bearing fault diagnosis. Characteristic parameters, such as crest factor, impulse factor and kurtosis, are sensitive to estimate abnormal noise.