基于支持向量机的铁水硅含量的预测  

Prediction of silicon content in hot metal based on support vector machine

作　　者：谈霖 张宗旺[1] 任冰朗 张海滨[2] 

机构地区：[1]北京科技大学冶金与生态工程学院,北京100083 [2]首钢股份公司迁安钢铁公司炼铁分厂,河北迁安064404

出　　处：《钢铁研究》2016年第1期6-9,共4页Research on Iron and Steel

摘　　要：为了准确预测高炉炼铁过程的硅含量,分析了高炉工艺参数对高炉铁水硅含量的时序性影响,以支持向量机理论为基础构建了2类铁水硅含量预测模型,即硅含量模型和硅变化量模型。利用首钢迁钢3号高炉铁水硅含量数据进行模型测试,测试结果表明2类模型预测命中率均可达到80%。To predict the silicon content in hot metal during BF ironmaking process accurately,the sequential influence of the BF process parameter on the silicon content in hot metal was analyzed.Two prediction models of the silicon content in hot metal were established based on support vector machine theory,which were silicon content model and silicon variation model.The models were tested with the silicon content data of No.3 BF in Shougang Qian＇an iron steel Co.Ltd..The results showed that the prediction hit rates of both models came up to 80%.

关 键 词：支持向量机 铁水含硅量 预测模型 

分 类 号：TF513[冶金工程—钢铁冶金]