Research On Aluminum Quality Prediction Modeling And Optimization Control Strategy Based On Chaos Grey Wolf Optimization

For the characteristic of bauxite in China,Bayer method,sintering method and mixing method have been adopted in the alumina production,and the sintering process in Bayer is the indispensable and last process in alumina production,which can have a very important impact on the quality.In view of the long process,the detection of large hysteresis,non-linear and complex process mechanism,serious relationship between equipment and fluctuations frequently on working condition,such as changes in raw materials,equipment scarring and other problems,which could result in that it is difficult to achieve automatic control and operation optimization by traditional strategy.Therefore,the importance and complexity of sintering process,as well as the modernization of the requirements with the industrial development,which makes that sintering process control system is paid more and more attention.In this paper,the research on the quality prediction in the sintering process and operation optimization under comprehensive conditions are discussed,which is based on an aluminum plant in Guangxi province.Meanwhile,some research results have been obtained.The main contents in this paper are summarized as follows:(1)Firstly,based on the analysis of the mechanism and characteristics in the sintering process,the factors that affect aluminum quality are found out.Then,the field data are filtered by the method of five points three times,and the outliers also eliminated by correlation analysis and box diagram analysis.(2)Secondly,a combination of principal component analysis and correlation analysis is used to select the auxiliary variables of the soft measurement model of alumina mass index as the model input variables.Then,chaos grey wolf optimization algorithm has been adopted to optimize the initial weights of online sequential limit learning machine in order to establish the alumina quality prediction model.The soft measurement model established in this paper is verified by the industrial field data,which has high precision and generalization ability.Achieving the online estimation of alumina quality indicators is subject to create the conditions for optimal control in aluminum sintering process.(3)Based on the above research work,currently,the present situation and existing problems of control parameters in the sintering process are analyzed further.Under the premise of ensuring the quality of alumina,comprehensive condition and operation optimization model in alumina sintering process has been established.Then,chaotic grey wolf optimization algorithm with constraints has been used to solve that model established and obtain the optimized operating parameters,which is main furnace temperature,ID fan power,feeding amount and so on.The simulation results show that the optimal control model is effective.(4)The operation optimization control system under comprehensive conditions for alumina roasting process is designed.Depending on the Ethernet and industrial cable,the system connect with the equipment layer and the control layer through internal protocol conversion to realize data interaction and the functional and information technology of control system.The experimental results show that the established optimized control system can obtain the optimal setting value of the basic control circuit according to the real-time change of the parameters in alumina production process,and the alumina quality index and its corresponding key operating parameters are controlled within the target range,and ensure the normal operation in alumina production process.