作者：张小乐¹, 于凯², 张胜男², 崔桂梅²

作者单位：1. 内蒙古安科安全生产检测检验有限公司，内蒙古 包头 014010;
2. 内蒙古科技大学信息工程学院，内蒙古 包头 014010

关键词：炉缸热状态;水温差;炉温;时间序列

摘要：

为准确预测炉缸热状态，考虑到炉缸热状态受高炉冷却壁段水温差影响而产生的时滞性，该文提出一种基于高炉水温差时间序列的新型炉缸热状态趋势评估预测方法。首先，采用相关性分析法计算出泊松相关系数并对历史炉温、与炉温预测相关的过程参数及水温差时间序列进行有效性提取；其次，综合高炉各项指标参数及相关历史信息，建立引入水温差时间序列的炉缸热状态评判预测模型，其中以PSO-LSSVM、BP-NN（BP神经网络）预测模型为例；最后，以某大型高炉为基础，预测分析其工业数据。预测模型的高命中率结果表明，引入高炉水温差时间序列方法可更准确地评估预测炉缸热状态。

Prediction method of blast furnace hearth thermal state by introducing time series of water temperature difference
ZHANG Xiaole¹, YU Kai², ZHANG Shengnan², CUI Guimei²
1. Inner Mongolia Anke Safety Production Testing and Inspection Co., Ltd., Baotou 014010, China;
2. School of Information Engineering, Inner Mongolia University of Science and Technology, Baotou 014010, China
Abstract: In order to predict the thermal state of hearth accurately and consider the time lag effect of the water temperature difference in the cooling stave sections of the bf hearth thermal state, a new trend prediction method for the thermal state based on the time series of the bf water temperature difference is proposed. Firstly, the correlation analysis method is used to calculate the Poisson correlation coefficient and extract the effective historical furnace temperature, water temperature difference time series and process parameters related to furnace temperature prediction. Secondly, the index parameters of the three-dimensional BF and their related historical information are comprehensively used to to establish a hearth thermal state prediction model by introducing the time series of water temperature difference, which takes the PSO-LSSVM and BP neural network (BP-NN) prediction models as examples. Finally, based on a large-scale blast furnace, the industrial data are predicted and analyzed.The high hitting rate of the two prediction models shows that the model introduced into the BF water temperature difference time series can more accurately predict the heat state of the hearth.
Keywords: hearth thermal state;water temperature difference;furnace temperature;time series
2022, 48(6):74-79  收稿日期: 2020-10-10;收到修改稿日期: 2020-12-21
基金项目: 国家自然科学基金（61763039）
作者简介: 张小乐（1981-），女，内蒙古包头市人，工程师，主要从事仪器仪表检测
参考文献
[1] 杨贵军, 蒋朝辉, 桂卫华. 基于熵权-可拓理论的高炉软熔带位置状态模糊综合评判方法[J]. 自动化学报, 2015, 41(1): 75-83
[2] 陈明, 尹怡欣, 祝乔. 基于分布参数模型的高炉炉温预测[J]. 控制理论与应用, 2014(9): 1232-1237
[3] 宋菁华, 杨春节, 周哲. 改进型 EMD-Elman 神经网络在铁水硅含量预测中的应用[J]. 化工学报, 2016, 67(3): 729-735
[4] GAO C H, JIAN L, LUO S H. Modeling of the thermal state change of blast furnace hearth with support vector machines[J]. IEEE Transactions on Industrial Electro- nics, 2012, 59(2): 1134-1145
[5] 宋贺达, 周平, 王宏. 高炉炼铁过程多元铁水质量非线性子空间建模及应用[J]. 自动化学报, 2016, 42(11): 1664-1679
[6] JIAO K X , ZHANG J L, LIU Z J. Analysis of the relationship between productivity and hearth wall temperature of a commercial blast furnace and model prediction[J]. Steel Research Interntional, 2017, 8(9): 1869-344
[7] 崔桂梅, 鄢常亮, 关英辉. 基于支持向量机的高炉向凉、向热炉况预测[J]. 钢铁研究学报, 2011, 23(7): 18-22
[8] 崔桂梅, 蒋召国, 詹万鹏. 基于时间序列的神经网络高炉炉温预测[J]. 冶金自动化, 2015, 39(5): 15-21
[9] 志成, 徐卫亚, 徐飞. 基于混合核函数PSO-LSSVM的边坡变形预测[J]. 岩土力学, 2012, 33(5): 1421-1426
[10] 孙斌, 姚海涛. 基于PSO优化LSSVM的短期风速预测[J]. 电力系统保护与控制, 2012, 40(05): 85-89
[11] CUI G M, JIANG Z G, LIU P L,et al. Prediction of blast furnace temperature based on multi-information fusion of image and data[C]//Proceedings of 2018 Chinese Automation Congress, 2018.
[12] 林旭梅, 刘帅, 石智梁. 基于改进PSO-FNN算法的钢筋混凝土腐蚀检测研究[J]. 中国测试, 2020, 46(12): 149-155
[13] 但长林, 李三雁, 张彬. 基于样本熵和SVM的滚动轴承故障诊断方法研究[J]. 中国测试, 2020, 46(11): 37-42