摘要
基于45 t转炉炼钢实际生产数据,通过数据预处理和互信息(MI)法进行特征选择,采用贝叶斯算法(BOA)优化BP神经网络模型、支持向量回归机(SVR)模型和LGBM模型的参数,预测转炉吹炼的耗氧量。用1176炉的实际生产数据对模型进行训练,504炉的数据用于验证模型的预测效果。结果表明,在预测的氧气体积偏差分别为±50、±40、±30 m^(3)的范围下,LGBM模型的预测命中率分别为94.04%、85.91%、76.58%。与SVR模型和BP神经网络模型相比较,LGBM模型有着更高的预测精度和稳定性以及更强的泛化能力。
Models of oxygen consumption in a 45 t converter were set up based on actual production data with BP neural network,support vector regression machine(SVR)and LGBM(Light Gradient Boosting Machine)algorithms where their parameters were optimized with Bayesian optimization algorithm and the models′features were selected by data preprocessing and mutual information methods.The actual production data of 1176 heats were used to train these models,and the data of another 504 heats were used to verify the prediction results of the models.The results showed that the hit rate with LGBM model within±50 m^(3),±40 m^(3)and±30 m^(3)was 94.04%,85.91%and 76.58%,respectively.Compared with the support vector machine and BP neural network models,the LGBM model had higher prediction accuracy and stability and better generalization ability.
作者
杨仕存
魏志强
钟良才
于学渊
李强
陈海娇
高威
YANG Shicun;WEI Zhiqiang;ZHONG Liangcai;YU Xueyuan;LI Qiang;CHEN Haijiao;GAO Wei(Low-Carbon Iron&Steel Frontier Technology Research Institute,Northeastern University,Shenyang 110819,China;School of Metallurgy,Northeastern University,Shenyang 110819,China;Steelmaking Plant of Fushun New Iron&Steel Company,Jianlong Group,Fushun 113001,China)
出处
《炼钢》
CAS
北大核心
2022年第4期7-13,共7页
Steelmaking
基金
中央高校基本科研业务费资助项目(N2125018)
国家自然科学基金项目(51574069)
科技部国家重点研发计划资助项目(2017YFB0304100)。
关键词
转炉炼钢
耗氧量
数据驱动模型
命中率
converter steelmaking
oxygen consumption
data drive model
hit rate