Font Size: a A A

Research On Application Of Fuzzy Neural Network In Temperature Control Of Rare Earth Electrolysis

Posted on:2022-10-14Degree:MasterType:Thesis
Country:ChinaCandidate:S BaiFull Text:PDF
GTID:2481306554467714Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Rare earths contain 17 categories of elements.Nd Fe B permanent magnet plays a very important role in new energy,electronic information and other high-tech industries.China is becoming the world's largest producer of Nd Fe B.Under the situation that the annual output of Nd Fe B continues to increase,the demand for neodymium metal as an important raw material for its addition has also begun to increase year by year.In the process of electrolytic neodymium oxide to produce neodymium metal,temperature affects the quality of neodymium metal electrolysis.Domestic rare earth industry develops relatively late,and the temperature control process of molten salt electrolytic neodymium oxide uses more manual empirical adjustment of the temperature in the tank.In addition to this backward production method,the quality of electrolytic products is unstable and the output is limited.Every year,while meeting the international market's demand for Nd Fe B materials,accompanied by huge waste of electric energy,the intelligent temperature control of the neodymium oxide electrolytic cell is imminent.During the investigation of a rare earth electrolysis workshop in Hezhou,Guangxi,the manufacturer pointed out that there were fluctuations in the quality of neodymium metal during the electrolysis of neodymium oxide.The production status of neodymium oxide fluctuates obviously.According to on-site observation and theoretical analysis,in the process of molten salt electrolysis of neodymium oxide,the temperature of the electrolytic cell is judged by the naked eye by manually observing the color of the molten salt,artificially changing the electrolysis current or the cathode insertion depth to adjust the temperature in the cell.The most critical influencing factor in the process,these backward electrolytic cell temperature control methods cannot accurately control the temperature in the cell,resulting in a series of unstable neodymium metal quality.The temperature control of the neodymium oxide electrolyzer is a multi-input,singleoutput system.The control process has time lag and nonlinear characteristics.Fuzzy neural network has good control ability to complex system,and the prediction ability of neural network can improve the system lag.This paper proposes to use fuzzy neural network algorithm to replace manual decision-making for temperature control.The fuzzy neural network has better control ability than PID algorithm.It is theoretically confirmed that the fuzzy neural network can be applied to the temperature control of rare earth electrolysis;The three factors of current,cathode insertion depth,and cathode current density respectively affect the experimental data of the electrolysis temperature.Use distance correlation to analyze the influence of the three factors on the temperature of the neodymium oxide electrolytic cell and determine the main variables of temperature control,and build the neodymium oxide electrolytic cell temperature control The experimental test platform tests the effect of fuzzy neural network algorithm on the temperature control of 6KA neodymium oxide electrolyzer.In the temperature control application of the 6KA neodymium oxide electrolyzer,the fuzzy neural network takes the cathode current density as the main control variable.It has a better control effect in the temperature control experiment.Compared with the traditional electrolytic temperature control method,the main control variable is uncertain and artificial.The temperature control and other reasons lead to unstable product quality.The temperature control method studied in this paper has has a better temperature control ability of electrolysis and a higher power usage rate,which largely stabilizes the quality of neodymium metal products.Rare earth electrolysis is an electrolysis process of various rare earth elements in electrolyzers of different structures according to the scale.Due to the differences in the structure and material of different specifications of neodymium oxide electrolysis cells,whether the research results of 6KA neodymium oxide electrolysis cell temperature control can be applied to other specifications of neodymium oxide electrolysis temperature control needs further study.
Keywords/Search Tags:Fuzzy neural network, PID, Cathode current density, Neodymium oxide electrolysis, Orthogonal temperature measurement
PDF Full Text Request
Related items