Multi-stage Noise Reduction Extraction of Bearing Fault Signal Features Based on Resonance Sparse Decomposition
Considering the characteristic frequency of bearing fault envelope spectrum and the periodic characteristics of frequency doubling components comprehensively,a judgment index that can eliminate the influence of accidental impact load and the periodic impact characteristics of strong background noise is established.For the requirement of RSSD method which needs to set the quality factor by manual method,standard ESMK is established as the fitness function.Then the fault signal of the bearing running process is diagnosed according to the envelope spectrum.The results show that the impact amplitude of the bearing fault is larger than that of the bearing fault when it is subjected to accidental impact,and the clear impact characteristics of the bearing outer ring fault cannot be obtained at this time.The double frequency data of 176 Hz and 263 Hz are formed,indicating that there is an outer ring fault in the bearing.The results of the inner ring fault signal show that no obvious inner ring fault characteristic frequency can be observed after the strong interference of noise factor,and the impact component has been significantly enhanced.An obvious frequency doubling of 261.1 Hz and 391.2 Hz is formed,indicating that there is an inner ring fault in the bearing.This research can effectively extract the bearing vibration signal characteristics,and can also be applied to other mechanical transmission fields.
NETL
NSTL
万方数据
