Research On Fault Diagnosis Of Elevator Guide Shoe Based On Multivariable Predictive Model

The elevator guide boots as one of the most important parts in the elevator structure,once the failure occurs,it will directly affect the comfort of the elevator and even the life.Therefore,the analysis of the state of the elevator guide boots and the safety detection are the problems that can not be ignored.When studying and analyzing the early failure state of elevator guide shoe,vibration signal analysis method can be adopted and combined with pattern recognition to detect it.In this paper,the vibration signal analysis method is used to decompose the complex signals obtained in the elevator guide boots,eliminate the chaotic signals and unrelated signals,get the abrupt signal related to the fault signal,then reconstruct the signal,decompose the signal after reconstruction,and get one single component signal,and take the energy value of the single component signal and pair the signal.It is normalized and is used as the input of the Variable predictive model based class discriminate(VPMCD)pattern recognition method.Therefore,this paper combines the time-frequency analysis method with the multivariable prediction model to study the fault diagnosis of the elevator guide boots.This article will conduct in-depth research on elevator guide shoe from the following aspects:1.Analyze the structure and vibration mechanism of the elevator,make a reasonable analysis of the failure form of the elevator guide boots,explore the relationship between the vibration acceleration signal of the elevator and the fault of the elevator guide boots,and explore and analyze the research object of the elevator's upper guide boots and the failure form of the lower guide boots.This study has important reference value for the safety of elevator.2.The principle of time-frequency analysis algorithm of empirical mode decomposition in nonlinear and non-stationary signal processing and the application of Hilbert transform in signal analysis are studied.The method of optimizing EMD by SVD is proposed.The method effectively improves the problem of extracting the early weak fault characteristics directly by EMD,and extracts the fault characteristic frequency of the vibration signal more accurately,and the validity of the method is verified.It3.The principle of VPMCD method,neural network and support vector machine is studied,and the three methods are compared and analyzed.From the experimental results,it can be seen that the multivariable prediction model is superior to the neural network and support vector machine,which shows the superiority and applicability of the VPMCD in the elevator guide boots.The SVD optimization EMD method and the VPMCD method are applied to the elevator guide boots,which provides an effective new way to realize the on-line intelligent diagnosis of the elevator boot fault.