Data Analysis Method Of Aluminium Cells Anod

With the large-scale development of the current in aluminum electrolytic cells, and constantly improve the energy saving requirements, the cell current distribution data accuracy and continuity has become more and more prominent. How electrolytic production process based on large amounts of data collected, using data analysis method to obtain accurate and reflect the current status of key data and information channel, which is the aluminum cell data analysis hotspot.This Paper takes500kA aluminum electrolytic cells as research object in some aluminum smelting factory. Describes the data collection method based on the anode current magnetic field and pressure drop equidistant designed anode field real-time monitoring system that can achieve continuous measurement anode magnetic data.In view of the actual condition of electrolytic cells, we used the COMSOL software to simulate currents and magnetic field data for subsequent analysis, and at the same time Gives the distribution of the magnetic field near the anode, analyzes the causes of complex magnetic field trip.The topic selected site collection magnetic field data on both sides of the anode guide rod of A and B as variables as variables in factor analysis. Based on a lot of experimental analysis, the results show that B side magnetic field data are much more convincing in electrolytic cells condition diagnosis. Next, we select the current data on both sides of the anode guide rod of A and B on the scene as the variables of the power spectrum estimation analysis, the analysis results are consistent with the results of magnetic field data analysis. On the above analysis, To eliminate unwanted noise data, we take different wavelet on the current data of anode guide rod B side de-noising processing with wavelet de-noising method, Extract the main features of the amount reflected trough conditions change.In addition, the task for these susceptible to noise and field-based data obtained equidistant drop the issue, discussed the anode temperature distribution by establishing temperature-current model approach, a curve-fitting of current and temperature data which are processed by wavelet de-noising method is given with the regression analysis to get the current formula changed with temperature of anode guide rod, we estimate the current distribution based temperature distribution as the basis of fault monitoring.