Research On Optimization Of Alumina Roasting Process For Equality Yield And Energy Consumption Index

As the main raw material for machinery,petrochemical,metallurgical and fertilizer industries,alumina has become widely used in aerospace,medical,automotive and semiconductor industries.For the alumina roasting process with strong nonlinearity,long flow and detection lag,it is difficult to realize modeling and optimization by traditional methods.An aluminum factory in Guangxi is taken as the study background,taking high production,high quality and low consumption as the optimization target,and focuses on the modeling and optimization of alumina roasting process for quality,production and energy consumption index,the following research results are obtained:(1)Firstly,the process mechanism of the roasting process and the connection between the integrated alumina production indicators,state parameters and operating parameters are analyzed to determine the optimal control objectives for the roasting process.The field data are pre-processed using multiple interpolation for missing values,K-means clustering algorithm to remove outliers,and data normalization process.(2)Secondly,the input variables for the alumina quality,yield and energy consumption prediction model are selected through process mechanism analysis combined with gray correlation analysis.The structural parameters(penalty factors and kernel width coefficients)of the Least Squares Twin Support Vector Regression Machine(LSTSVR)are optimized using the Quantum Chaos Salp Swarm Algorithm(CQSSA)to develop alumina quality,yield and energy consumption prediction models.The simulation experiments are verified by using the actual data of industrial production,the results show that the soft measurement model developed in this paper is able to achieve effective estimation of alumina production indexes,meet the actual industrial soft measurement requirements and provide the prerequisite for optimal control of the roasting process.(3)Thirdly,in order to achieve the optimization objectives of high production,high quality and low consumption,an optimization model of operating parameters with maximum alumina production and minimum energy consumption as the optimization objectives and alumina quality as the constraint is established on the basis of the prediction model.The model is solved using a Multi-objective Salp Swarm Algorithm with constraints to obtain the optimized operating parameters: fan power,gas flow rate,fan speed,and so on.The simulation is verified using real production data from the plant.The simulation results show that the optimization strategy can ensure stable operation of the alumina roasting process and achieve optimized operation of the roasting process to achieve the production target of high yield,high quality and low consumption.