Design And Research Of Alumina Concentration Control System

The control of alumina concentration has always been the key to the production process of electrolytic aluminum.This paper takes the control of alumina concentration in the process of electrolytic aluminum production of an Electric Aluminum Company as the research background,and adopts intelligent feature modeling theory,single neuron adaptive control and Neural network soft measurement and other methods are applied to the control system of alumina concentration,which achieves certain control effects.There are two basic methods for controlling the concentration of alumina: one is to construct a mathematical model in the design to achieve precise control,but the complexity of the production process makes it difficult to establish an accurate mathematical model to control the concentration of alumina.The second is the intelligent control technology,which will be used as the main way to achieve the control of alumina concentration.However,it is difficult to achieve high-efficiency control due to many unfavorable factors such as high temperature,high current and harsh environment.The intelligent feature modeling theory designed by this subject establishes a model that combines the advantages of the above two control methods,and can comprehensively integrate fuzzy and accurate models,qualitative and quantitative models,thereby improving control precision and optimizing the control of alumina concentration..Based on intelligent control theory,this topic analyzes the relationship between alumina concentration and apparent resistance through intelligent models,single neuron adaptive PID control,Elman neural network,least squares and other related theories.According to the data obtained from the analysis,an intelligent model of alumina concentration is constructed.The model includes the control model between the feed and the apparent resistance,the model of the working area,the characteristic model of apparent resistance and alumina concentration,the qualitative estimation model,and the integrated integration.Models,etc.According to the model,the whole control system was designed and simulated.Compared with the original method,this method is better than the previous method in controlling the concentration of alumina.For the whole system,due to many uncertain characteristics of the controlled object itself,it is necessary to identify the control model parameters online in real time,so the workload of the control system is increased,and the real-time performance of the control system is also reduced.Later,more time will be devoted to in-depth study of these issues.