Research On Rolling Bearing Fault Feature Enhancement Method Combining Filtering Algorithm And Quantum Theory

Rolling bearings are key components in modern mechanical equipment and are widely used in industry.Once a rolling bearing fails,it may cause a safety accident or even casualty.In order to detect the bearing status in real time without disassembling the machinery and equipment,sensors are usually used to monitor the vibration signal of the bearing,and noise suppression and signal enhancement are performed by specific algorithms to achieve the feature extraction of bearing failure.This paper introduces the correlation analysis method to pre-process the noise-reduced signal fragments,and conducts a study on the bearing vibration signal noise reduction algorithm based on the non-local mean noise reduction algorithm and bilateral filtering algorithm,and conducts a study on the signal enhancement algorithm based on the quantum theory and non-local mean algorithm.The research content of the thesis mainly includes:(1)A method for matching the time domain features of bearing vibration signals.In order to integrate the signal fragments of the same moment of different periods,this paper splits the one-dimensional vibration signal into several fragments,proposes to use the correlation coefficient as the basis for selecting the similarity measure,and integrates the signal fragments to be reduced and all the similar fragments to obtain the window matrix,which not only lays the foundation for the subsequent noise reduction,but also is an important guarantee for enhancing the fault information inside the vibration signal.(2)Research on vibration signal noise reduction methods based on non-local mean and bilateral filtering.Based on the analysis of the core ideas of the non-local mean noise reduction algorithm and the bilateral filtering noise reduction algorithm in the field of image noise reduction,the core ideas of the above two algorithms are applied to the window matrix in combination with the characteristics of the window matrix obtained in the article,and the improved algorithms are analyzed and applied.In addition,combining the advantages of the non-local mean and bilateral filtering algorithms,a new noise reduction algorithm combining the non-local mean noise reduction algorithm and reconstructing the bilateral filtering is proposed.(3)Research on vibration signal noise reduction methods based on non-local mean and quantum theory.This paper first analyses the reasons why the non-local mean noise reduction algorithm can obtain excellent noise reduction capability in the background without impulse noise.This is followed by describing the degree of confusion between the signal fragment to be reduced and similar signal fragments by means of the superposition state in quantum theory,and the impulse noise is precisely removed based on the difference between the signal point to be reduced and the corresponding similar point in the degree of confusion.Finally,noise reduction is performed by non-local mean.(4)Research on the vibration signal noise reduction method based on quantum theory reconstruction of non-local mean.A quantum description of the bearing vibration signal is proposed.Based on the non-local mean noise reduction algorithm,a multi-quantum bit system is established for the time-domain vibration signal,which combined with the window matrix can enhance the fault information better than other algorithms described in the paper.The above research results are validated with actual bearing data,and the experimental results show that the approach of this topic not only achieves good noise reduction,but also has a significant effect on the enhancement of the eigenfrequencies in the Hilbert envelope spectrum.In addition,this topic enriches the research of quantum theory in the field of noise reduction and in the field of bearing fault feature extraction.