Normal cell voltage data time frequency analysis method and device used for aluminum electrolysis cell process control

Abstract

The invention discloses a normal cell voltage data time frequency analysis method and device used for aluminum electrolysis cell process control. The method comprises the steps of performing time-frequency analysis on historical data of a normal cell voltage; and applying energy spectrum density based on a scalogram to obtain a frequency range corresponding to aluminum oxide concentration and low-frequency cell noise frequency division and frequency band widths related to cell stability. By implementing the method, the deficiencies and defects in the prior art can be overcome, and an obtainednormal cell voltage comprises multiple signal components and corresponding frequency ranges; compared with the prior art, related frequency ranges are narrower, distribution of related signal components is more accurate, and sub-low-frequency noises caused by other factors not involved in the prior art are also covered.

Description

technical field [0001] The invention relates to the field of aluminum electrolytic cell process control, in particular to a time-frequency analysis method and equipment for normalizing cell voltage data used for aluminum electrolytic cell process control. Background technique [0002] Cryolite-alumina molten salt electrolysis is the only way for the aluminum electrolysis industry to produce primary aluminum, and the aluminum electrolysis cell is the "heart" of the aluminum electrolysis production process. The raw material used in electrolysis is alumina, the electrolyte is molten cryolite, and a carbon anode is used. The result of electrolysis is molten aluminum on the cathode and carbon dioxide on the anode. . Due to the complex electrochemical, physical and chemical reactions involved in the aluminum electrolytic cell, the cryolite-alumina molten salt has the characteristics of high temperature and strong corrosion, so many key parameters cannot be directly measured onli...

AI Technical Summary

Problems solved by technology

Therefore, the use of spectrum analysis based on Fourier transform cannot analyze the frequency composition of the tank voltage signal and its variation with time, and cannot clearly distinguish the corresponding relationship between the components of the tank voltage signal and their physical meanings; using Fourier transform Spectrum analysis of non-stationary signals is also prone to false spectral peaks, which in turn affects the accuracy of signal component acquisition related to alumina concentration state information and tank noise[2]

[0009] In terms of obtaining alumina concentration state information and groove noise signal components, due to the limited ability of inertial filters or Kalman filters to deal with non-stationary signals, the high-pass, band-pass and The low-pass filter cannot separate the information components of each frequency band in the tank voltage signal well, and produces delays, which in turn affects the accuracy and real-time performance of key control parameters for alumina concentration control and tank stability

[0010] In summary, the existing cell voltage analysis method based on the combination of spectrum analysis and multi-filter cascading can obtain the alumina concentration state information and the corresponding frequency band of the cell noise, so that the separated from the cell voltage signal and alumina Concentration-related state information still contains a lot of low-frequency groove noise, and there is a lot of subjectivity in the selection of parameters in the signal component separation method designed based on the inertial filter or Kalman filter and other filter cascades, which affects the oxidation. The accuracy and real-time performance of key parameters in aluminum concentration control and tank stability control, which in turn affects the performance of the tank control machine control system

Images