The Optimization Model Of Burden Surface In Blast Furnace For Production Indexes

Iron and steel industry is an important pillar industry of the national economy.It is an important indicator for measuring a country's industrial level and comprehensive national strength,and it has non-negligible impacts on sustainable development.Blast furnace(BF)ironmaking process is the dominant unit for producing the molten iron in the metallurgical industry,whose energy consumption accounts for the main part of the whole process."Made in China 2025”plan advocates"integration of informationzation and industrialization"as the main line,strengthens industrial basic capabilities,improves industrial techniques and product quality,and accelerates the development of energy efficiency and technological innovation.Therefore,the realization of informationization and intelligentize for BF ironmaking meets the needs of national development.In the four operating systems of BF ironmaking process,the upper regulation system(i.e.,charging system)is one of the most frequent and effective means to control the production state of BF,which directly affects the burden distribution at the furnace throat.Reasonable burden distribution can make the gas flow distribution more reasonable and the chemical reaction in the furnace more fully.It plays a vital role in the long-term stable operation,and energy-saving and emission-reduction.Due to the complicated operating mechanism and the difficulty In detecting key parameters,burden surfaces are usually tuned by experienced operators.Such kind of manual tuning cannot ensure the key production indexes to be within the target ranges.Therefore,a data-driven optimization scheme is proposed for achieveing the optimal setting of burden furnace in this thesis,which carries out a study on BF production indexes system,production indexes prediction model,optimization model of burden surface and feedback compensation model.The main contents and achievements are Iisted as follows:(1)The production indexes system is constructed and the optimization scheme for deterrmining the setting values of burden surface is proposed.Firstly,according to the production purpose and practical experience of BF.a feasible production indexes system is constructed from five aspects:quality index.cost index,energy utilization index,smooth index and operating parameters index.In addition.the multi-scale definition and characterization of these production indexes are carried out to guide the direction of subsequent research.Then.considering the difficulity of burden surface optimization,a multi-level optimization scheme is proposed for determining the setting values of burden surface based on data-driven technique,which combines intitial setting model of burden surface,production indexes prediction model and feedback compensation model togther.(2)The production indexes prediction model is established.The nonlinear relationship model among the production indexes,burden surface parameters and operating parameters are estsblished based on extreme learning machine(ELM)theory.According to the characteristics of BF data,a variety of improved ELM algorithms are proposed to enhance the application ability of the algorithm to process industrial data and improve the accuracy of the model.The real production data collected from BFs are applied and validated the reliability and effectiveness of the proposed algorithms.Comprehensive experiments indicate that the proposed algorithms can achieve better performance.(3)The multi-objective optimization model for determing the setting values of burden surface is proposed.In order to achieve the goal of production optimization with high efficiency and low consumption,this multi-objective optimization model takes the gas utilization ratio and coke ratio as optimization objectives,production smooth and product quality as constraints,and the burden surface features as decision variables.Then,an integrated multi-objective optimization solution framework is proposed based on the characteristics of the optimization problem to find feasible solutions to obtain the more comprehensive Pareto optimal solutions.(4)A feedback compensation strategy based on association rule mining is presented.Firstly,an incremental rule form between production indexes and burden surface features is constructed.Then,the compensation rules are extracted from a large number of data produced during the BF ironmaking process by the improved association rules mining method.When the optimized values deviates from the actual values,the rules are used to find the corrected values to further compensate the initial setting values of burden surface.The simulation experiments are carried out to verify the feasibility and validity of the proposed method.The results demonstrate that the proposed optimization scheme can give a reasonable burden surface so as to make the production indexes whthin their target ranges and provide a basis for subsequent charging.