基于柜位预测的转炉煤气热值优化模型  

Optimization model of converter gas calorific value based on cabinet location prediction

作　　者：陈光 胡鹏 包向军 郦秀萍 刘骁[3] 张璐 CHEN Guang;HU Peng;BAO Xiangjun;LI Xiuping;LIU Xiao;ZHANG Lu(School of Energy and Environment,Anhui University of Technology,Maanshan 243000,Anhui,China;Steel Industry Green and Intelligent Manufacturing Technology Center,China Iron and Steel Research Institute Group Co.,Ltd.,Beijing 100081,China;Central Iron and Steel Research Institute Co.,Ltd.,Beijing 100081,China)

机构地区：[1]安徽工业大学能源与环境学院,安徽马鞍山243000 [2]中国钢研科技集团有限公司钢铁绿色智能化中心,北京100081 [3]钢铁研究总院有限公司,北京100081

出　　处：《钢铁》2024年第3期197-206,共10页Iron and Steel

基　　金：国家重点研发计划资助项目(2020YFB1711101);安徽高校自然科学研究资助项目(KJ2021A0411)。

摘　　要：转炉煤气是转炉冶炼时产出的重要能源介质,提升转炉煤气热值对用户使用及平衡调度等有着重要意义。目前大多数研究都针对提升回收煤气量,而对通过调控提升煤气回收质量研究未见报道。为了在柜位回收限度内提升转炉煤气的回收质量,增加转炉煤气余热余能回收效率,结合柜位预测调控起止回收时刻,建立了转炉煤气热值优化模型。通过煤气发生量与煤气消耗量的特征分析,依据吹炼计划建立碳平衡进而开发煤气平均发生流量预测模型,与实际生产数据相比,模型精度达96%。使用Sarima模型对历史数据训练并开发煤气消耗量预测模型,模型精度达97%,结合上述模型根据吹炼开始时初始柜位建立了煤气柜位预测模型,预测吹炼周期柜位变化规律,模型精度达95%。根据历史数据拟合出CO体积分数特征曲线,方差为0.95以上。利用非线性规划优化算法,以回收煤气热值为优化目标,柜容和起止回收CO体积分数为约束条件,开发出转炉煤气起止回收时刻调控模型并编程求解得到了优化方案,提升了转炉煤气回收热值,以某钢单炉调控结果为例,调控前后回收煤气热值从6278.3 kJ/m^(3)提高到6654.6 kJ/m^(3),并降低了高热值煤气放散。通过和大量现场数据对比,平均回收煤气热值提高5%,该建模过程为转炉煤气热值提升提供了有效的优化方法。Converter gas is an important energy medium produced during converter smelting,and improving the calorific value of converter gas is of great significance for user use and balanced scheduling.At present,most studies focus on increasing the amount of recovered gas,but there are no reports on improving the quality of gas recovery through regulation.In order to improve the quality of converter gas recovery within the limit of cabinet recycling,and increase the efficiency of converter gas waste heat and energy recovery.This article establishes an optimization model for the calorific value of converter gas by combining the prediction and control of the starting and ending recovery time of the cabinet.By analyzing the characteristics of gas generation and gas consumption,a carbon balance was established based on the blowing plan to develop a prediction model for the average gas flow rate.Compared with actual production data,the accuracy of the model reached 96%.The Sarima model was used to train and develop a gas consumption prediction model based on historical data,with an accuracy of 97%.Based on the initial cabinet position at the beginning of the blowing process,a gas cabinet position prediction model was established to predict the changes in cabinet positions during the blowing cycle,with an accuracy of 95%.Fit the CO concentration characteristic curve based on historical data,with a variance of over 0.95.Using nonlinear programming optimization algorithms,with the goal of recovering the calorific value of gas as the optimization objective,and the container capacity and CO concentration at the beginning and end of the recovery as constraints,a control model for the start and end of the recovery time of converter gas was developed.The optimization scheme was solved through programming,which improved the calorific value of converter gas during the blowing cycle.Taking the control results of a single steel furnace as an example,the calorific value of the recovered gas before and after the control was increased from 6

关 键 词：转炉煤气热值 柜位预测 非线性规划 CO体积分数 优化模型 

分 类 号：TF71[冶金工程—钢铁冶金]