Simulation Of Electrolytic Tank Voltage Temperature Signal Feature Extraction

Industrial electrolytic process is a multivariable time-varying, nonlimnear and strong coupling complex industrial system. The whole process of electrolysis are conducted in electrolytic cell, The state of the cell will changes constantly with the electrolytic process, both voltage and temperature signals are important parameters to reflect cell status accurately, So, Feature extraction for these signals is rather significant and valuable.This paper made the spectrum analysis of the laboratory simulation electrolytic cell, The Hilbert-Huang is used to feature extraction of cell voltage and temperature signal, the process is carried out in a PC. The Hilbert-Huang transform completely out of the Fourier analysis framework that suits for dealing with non-linear and non-stationary signals. The expected finding is the graded separation of the original signal based on local time scale, analyzing signal component, getting the signal frequency characteristics.In this paper by designing the experiment, collecting cell voltage and temperature signal under the different cell condition, and using HHT to process all signals collected. Through empirical mode decomposition and the Hilbert spectrum analysis, getting a series of different characteristic of intrinsic mode functions, Comparing the component frequency between different cell condition; On the whole period of the Hilbert integral to get the Hilbert marginal spectrum, Comparing the signal frequency energy distribution ender different cell condition. Through experimental analysis, the fluctuation of electrolyte, electric tank auxiliary heater working condition on the signal frequency characteristics, to prepare for tank condition recognition diagnosis system.The Hilbert-Huang Transform is applied for anode voltage signal processing of aluminum electrolysis, Simulation experiments showed the Hilbert-Huang Transform can accurately describe the anode voltage signal frequency component and the characteristics of the signal frequency, having great value in practical application and deserving further research.Tools in the research including Visual C++6.0, Sql Server2008, Matlab. Visual C++6.0realize serial data communication, data display, database connection, and other functions; Sql Server2008database complete data management; Matlab realize the function of signal processing and analysis through calling the data of database.