高炉铁水含硅量的动态神经网络多步预报  

Multi-step prediction of molten iron silicon content in blast furnace using dynamic neural network

作　　者：王玉涛[1] 严其艳[1] 杨钢[1] 徐万仁[2] 

机构地区：[1]流程工业综合自动化教育部重点实验室,沈阳110004 [2]宝山钢铁股份有限公司炼铁厂,上海200941

出　　处：《仪器仪表学报》2006年第11期1448-1451,共4页Chinese Journal of Scientific Instrument

基　　金：辽宁省自然科学基金(20062021)资助项目。

摘　　要：本文提出一种基于主成分分析法的动态神经网络模型实现高炉铁水含硅量多步预报。本方法将经主成分分析法处理后的高炉数据样本作为改进型Elman网络的输入,减少变量个数,简化网络结构；同时采用动态递归算法进行高炉铁水含硅量多步预报,具有较强的适应性。采用该方法对宝钢某高炉现场采集的数据进行预报实验,以±0．05作为预报误差,预报命中率达到89．12%。Based on principal component analysis （PCA）, a multi-step method to predict the molten iron silicon content using dynamic neural network is introduced. The data sample processed with PCA is used as the input of improved Elman network, so that the number of input variables is reduced and the network structure is simplified. Meanwhile, the dynamic recurrent algorithm is adopted to predict the molten iron silicon content in multi-step, which has strong adaptability. The method was used in a prediction experiment with the data collected from a blast furnace of Baosteel on-site. With the prediction error range of ±0. 05, the prediction hit rate reaches 89.12%.

关 键 词：铁水含硅量预报 主成分分析法 改进型Elman网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]