Font Size: a A A

Research On Vibration Fault Diagnosis Of Rotating Machinery

Posted on:2020-01-30Degree:MasterType:Thesis
Country:ChinaCandidate:Y Y HeFull Text:PDF
GTID:2392330578470242Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
With the development of rotating machinery in terms of large-scale,complicated and automated,people's requirements for efficiency,safety and reliability are increasing.Condition monitoring and fault diagnosis at the beginning of rotating machinery have an important impact on extending equipment life,improving equipment utilization,increasing equipment safety,reducing maintenance time and reducing maintenance costs.Aiming at the problem of extracting early fault characteristic information of rotating machinery,the paper uses two kinds of fault diagnosis methods based on kurtosis and VMD/EWT for nonlinear and non-stationary vibration signals with low SNR.The specific research contents of this paper are as follows:1)According to the simulation signal,this paper analyzes the lack of complete theoretical support,envelope fitting,end effect,and modal aliasing in the EMD method.On this basis,this paper analyzes the improved methods based on EMD,such as LMD,EEMD,and this paper uses the AM-FM simulation signal to compare the decomposition performance and limitations of the three methods.This analysis is a comprehensive and systematic summary analysis.2)Aiming at the parameter optimization problem of VMD method,this paper proposes the parameter K?????value strategy in principle,which solves the problem that the traditional debugging method has large workload,low efficiency and lack of reliability.The value of this variable can guide the actual application of VMD;3)This paper proposes a kurtosis-based VMD/EWT fault diagnosis method.Firstly,the original signal is decomposed by VMD/EWT,which can effectively separate the fault information of each frequency band.Secondly,use the kurtosis index to select the modal function that best reflects the fault characteristics,so as to avoid the influence of strong background noise.Finally,the composite signal is subjected to envelope analysis to achieve fault diagnosis.4)Based on the simulation and measured data,the fault diagnosis methods based on VMD,EWT and EMD are compared and analyzed from the aspects of decomposition mechanism,filtering method,decomposition layer and deco1position speed.In addition,the feasibility is verified.
Keywords/Search Tags:signal processing, fault diagnosis, Empirical Mode Decomposition, Variational Mode Decomposition, Empirical Wavelet Transform
PDF Full Text Request
Related items