Research On Iron Ore Sintering Optimization Method And Engineering Realization

The technology of iron ore agglomeration plays an important role in the smooth and stable production of blast furnaces and the cost reduction and efficiency enhancement of enterprises.In recent years,China's iron and steel enterprises have paid more attention to the research of sinter blending technology and its application in practice.However,at present,sinter distribution is determined mainly based on artificial experience,which takes the cost of raw materials of sinter more into consideration,while there is no qualitative analysis of the properties of sinter.A method is urgently needed to improve the quality of the sinter and the profits of the sintering plant while ensuring the lowest cost.This study is based on a sintering plant.Through the analysis of the process of ore sintering in a sintering plant,the mathematical model of ore sintering was established according to the actual production requirements.The model is solved by using the non-dominant sequencing genetic algorithm(NSGA-II)with elite strategy,and the optimal scheme of sintering distribution is obtained.The software platform of the upper computer is designed and developed.The main contents of this paper are as follows:Firstly,the entire process of ore sintering is analyzed.Because iron ore sintering is a complex physical and chemical reaction process,and the mechanism model is difficult to predict accurately,so mechanism model is adopted in combination with data drive.The state and parameter variables involved in the sintering process were analyzed,the mechanism model and data driven model input and output were determined,the mechanism model corresponding to the properties of mixed ore and sinter was established,and the relationship between the properties of sinter and the first-grade product rate and qualified product rate was fitted by SVR support vector regression.Then,starting with the problem of multi-objective optimization,this research analyzes the basic concept and common methods of multi-objective optimization,and the constraint conditions and objective functions were determined based on the actual situation in the field.After the genetic algorithm has been studied in detail,an improved genetic algorithm,noninferior sorting genetic algorithm with elite strategy(NSGA-II),was introduced,and the optimization model of sintering blending was solved.After that,the research applies the TOPSIS algorithm to the obtained non-inferior solutions,the optimal operation plan of the ?ore blending process was obtained.Finally,combined with the needs of industrial sites,the research sets up an optimized software platform for the Ore Sintering Optimization,and provides a human-computer interface that is easy for operators to use.This part specifically includes the preparation of server-side and client programs,database construction and association of the algorithm.Optimization software is applied to the actual industrial scene.After the operating plan obtained from the optimization algorithm is compared with the operating plan obtained from the manual experience in the field,the results of this study are validated,at the same time this research brings economic benefits for the enterprise.