| In recent years,iron and steel enterprises have serious overcapacity,and its profits continue to lower.The whole iron and steel enterprises are facing the main problem of how to reduce the production cost and raise efficiency.Using the low-grade iron ore is undoubtedly the effective measures to lower costs,but using the low-grade iron ore to smelt will bring a variety of problems.So,it is very necessary to study the BF ore-matching deeply.To the different ore grade,there will be the different ratio of slag and iron,and then the fuel ratio and pig iron costs are also different.It is necessary to consider the chemical compositions and metallurgical properties of the pellets and the natural lump ore,and so on.It also meet the requirements of coke performance,heat system,blast system,slag system and harmful elements into the furnace in the production of BF.What is more,it needs the lowest cost of pig iron.The research on the optimization of raw materials and ore-blending schemes at home and abroad mainly concentrates on the sintering and pellet,and the research on the BF ore-matching is relatively few.Based on the Big Data theory and the detailed production data from August 2009 to March 2015 for nearly 5 years of No.8th BF in Wisco,the artificial neural network model is established to train the blast furnace utilization coefficient and synthetic coke ratio,and then forecast them in this paper.The weights of the best experimental group were selected to calculate the influence coefficient of ore grade change on coke ratio,integrated into the optimization of the decision support system of BF distribution.According to the different silicon content of pig iron,wind temperature,iron grade of ore and Feo content of ore,the change of coke batch and fuel quantity is calculated respectively,so as to put forward the corresponding reasonable variable material scheme.Based on changing material scheme,the regulation of heat system and slag making system is carried out. |
PDF Full Text Request