Diagnosis Of The Cathode Condition Of 300KA Prebaked Anode Aluminum Reduction Cell Based On Wavelet Analysis

The process of industrial aluminum electrolysis is completed in the electrolytic cell,and the running state of the electrolytic cell directly affects the current efficiency and energy consumption,so it is very important to master the operation state of the cell.The operation status of aluminum reduction cell is related to many factors,and the voltage at the soft bus bar of the cathode of aluminum electrolytic cell is one of the important parameters to reflect the condition of the cell.Therefore,it is of great significance to accurately diagnose the cathode condition by analyzing the cathode soft bus voltage of aluminum reduction cell.The wavelet transform is used to analyze the voltage signal of the cathode soft bus of aluminum electrolysis cell.The voltage signals are analyzed from two aspects of time and frequency domain.The method of energy contrast experiment is used to determine the wavelet decomposition level.After making wavelet multi-scale decomposition to the voltage signal,the third layer approximation signal with less noise and more principal components is obtained,In the time domain,the characteristic parameter is obtained by using the method of dimensionless feature parameter extraction to the third layer approximation signal.Then the K-mean clustering analysis method is used to the characteristic parameters.And getting a class of data with a fault and a class of failures.However,the applicability of this diagnostic method is verified when the diagnostic time interval is reduced,and fault recognition rate can not meet the actual needs.In the frequency domain,the envelope spectrum of the third approximation signal is used to extract the characteristic parameter which is maximum value of the envelope spectrum.Through the statistical analysis of the maximum value of the envelope spectrum,the critical value of the fault is obtained,and the data are classified by the critical value.Then the data is divided into a class of failures and failure free.This method has passed the validation of applicability.Finally,the advantages and disadvantages of two kinds of diagnostic methods in time domain and frequency domain are compared.It is found that the diagnosis method of the classification analysis of the maximum value of signal envelope spectrum in frequency domain is better.So,the fault diagnosis method is applied to the design of cathode condition diagnosis system for aluminum electrolysis cell.The diagnosis system uses MySQL to store the collected data which is called by Matlab.Matlab achieves the function of calling for the data in the MySQL,viewing data functions,drawing curve function,diagnosis function.