Font Size: a A A

Research On Fault Diagnosis Method Of Roller

Posted on:2013-04-02Degree:MasterType:Thesis
Country:ChinaCandidate:X WuFull Text:PDF
GTID:2231330392454669Subject:Mechanical design and theory
Abstract/Summary:PDF Full Text Request
With the progress of the research of nonlinear dynamics, chaos appears theory, it isthe study of the nonlinear behavior of major breakthrough.Chaos determine macroscopicnonlinear systems under certain conditions rendered uncertain or unpredictable seeminglyrandom phenomena.Crack of roll in rolling, nonlinear phenomenon is very strong,therefore, In this paper, the chaos theory into the crack fault diagnosis.First introduced thedefinition and characteristics of chaos, To effectively extract the useful signal,a waveletdenoising research,preparation of a wavelet denoising procedures Matlab, the signal noise,and achieved good results.Then introduced the phase space reconstruction theory,and usethe C-C method obtain the Embedding dimension m and delay timeτ.Through the analysis of normal and crack condition roller rolling motioncharacteristics, solving the time series of chaotic vibration, The largest Lyapunov exponent,correlation dimension, Kolmogorov entropy characteristic parameter, And analysis ofnormal and characteristic parameters of cracks and crack different state, from anotherperspective to explore the new method of mechanical fault diagnosis, can lead to chaoticcharacteristics of volume changes on roll fault diagnosis.Time series forecasting is topredict things in the past and present observational data structure change by the time seriesmodel, by certain rules to speculate about the future, is the premise to predict themechanical state of motion.The application of chaos in time series prediction to manypreviously regarded as random signal phenomenon, now can use chaos certainty to explainit. Using phase space reconstruction theory, and based on echo state network for timeseries prediction, and based on the time series prediction research, for the operation of theroller status in predicting lays a foundation, enables us to predict the state of roller failureand fault degree to make a judgement, it has very strong actual application space, advancepreventive measures to improve the rolling mill equipment safety, raise economic benefitsto have important practical value.
Keywords/Search Tags:chaos, roll, lyapunov exponent, correlation dimension, kolmogorov entropy, echo state network
PDF Full Text Request
Related items