Intelligent Modeling Of Alumina Roasting Process Based On ELM And GA And Its Control System Research

In the past more than and 10 years,China's nonferrous metals industry has developed rapidly and made great contributions to the development of economic and social and national defense science and technology industry.As a raw material for producing aluminum,alumina plays an important role in aluminum smelting industry.Bayer is one of the main methods used in alumina production in China,and alumina roasting is one of the influential sectors of alumina quality,energy consumption and production cost.Using intelligent method to model the roasting process and the appropriate algorithm to study the optimization and control to the parameters of the roasting process is a direction of industrial innovation in alumina production industry and an effective way to improve the quality of alumina.Based on the technology of G.S.C,extreme learning machine(ELM)which optimized by improved particle swarm optimization(PSO)algorithm is used to establish the prediction model of alumina roasting process,and genetic algorithm(GA)is used to optimize the parameters of alumina roasting process.Design a process control system based on DCS for alumina roasting,and realize the precise control of key parameters in roasting process by BP_PID controller.Main contents are:(1)In order to solve the difficult problem of roasting process modeling,the prediction model of roasting temperature is established by BPNN,standard ELM and ELM optimized by improved PSO.After comparison,find that the ELM optimized by improved PSO prediction model that the coefficient is up to 0.9856 has obvious advantages than BPNN and standard ELM in prediction accuracy and generalization performance.(2)Aiming at the problem that the parameters of the roasting process are serious and the fluctuation of the working conditions is frequent,the optimization model of working condition is established by GA.The stable value of roasting temperature(1070?)under the condition of normal production for the control target,find the optimal combination of the parameters which have a great influence on the roasting temperature.Establish optimal working condition database based on this,and according to the monitoring of the temperature and the deviation between the set value,the control system find the optimal combination of working condition in the database to guide the real-time adjustment of the corresponding control variables which will make the production process in the best condition.So,it can avoid the subjectivity of artificial setting and the disoperation of production process,reduce unnecessary energy consumption,stable roasting temperature and Improve alumina quality.(3)In view of the lack of the level of automation and the imperfect production and management of alumina roasting process,the process control system of alumina roasting based on DCS is designed.Use BP_PID controller to realize process operation parameters control and process control system to realize the production process monitoring.Reduce production costs by reasonable allocation of production materials and improving production efficiency.(4)Taking the control of roasting temperature as an example,establish the object model,virtual execution mechanism and basic control loop on the advanced process control system simulation platform.The results of the simulation experiment show that the system can track the setting value of roasting temperature which can prove the feasibility of the control system.