高炉炉顶红外视频图像的智能化处理技术研究
|
摘要
提出一种基于高炉红外图像的煤气流中心区域分类以及煤气流中心点动态分布的方法,得出煤气流中心区域的8个发展阶段以及煤气流中心变化轨迹.将每小时内8个阶段的时间占比与煤气利用率结合进行FCM聚类,定义煤气流发展的3个阶段,分别为"初期阶段、中期阶段、旺盛阶段".聚类分析得出:当初期、中期和旺盛3个阶段的时间占比分别达到13. 85%,29. 88%和56. 27%时,煤气利用率最大,且煤气流旺盛阶段的时间占比与煤气利用率呈负相关关系.研究表明:煤气流中心点产生于距高炉物理中心西南位置椭圆的第四圈,其变化轨迹为:在西南位置与物理中心位置之间做往复运动.本文提出的方法能实时的了解高炉生产过程中煤气流分布的动态变化,进而实现煤气利用率的实时监测和调控.
According to the problem that the dynamic change of gas flow in blast furnace is difficult to identify accurately,a method was presented for the classification of gas flow central area and the dynamic distribution of gas flow center point based on the infrared image of blast furnace. The eight development stages of the central gas flow area and the trajectory of the gas flow center were obtained. FCM clustering was combined with the proportion of time in eight phases per hour and gas utilization. The three stages of gas flow development were defined as the followings: initial stage,medium stage,and vigorous stage. The cluster analysis showed that when the proportions of the three phases of the initial,medium,and vigorous periods reached 13. 85%,29. 88%,and 56. 27%,respectively,the gas utilization rate was the largest,and the proportion of time during the vigorous gas flow was negatively correlated with the gas utilization rate. Further research shows that the gas flow center point was generated from the fourth circle elliptical from the southwest position of the physical center of the blast furnace,and its changing trajectory is a reciprocating motion between the southwest position and the physical center. The method proposed herein can help understand the dynamic changes of gas flow distribution in the blast furnace production process in real time,and then realize real-time monitoring and regulation of gas utilization rate.
According to the problem that the dynamic change of gas flow in blast furnace is difficult to identify accurately,a method was presented for the classification of gas flow central area and the dynamic distribution of gas flow center point based on the infrared image of blast furnace. The eight development stages of the central gas flow area and the trajectory of the gas flow center were obtained. FCM clustering was combined with the proportion of time in eight phases per hour and gas utilization. The three stages of gas flow development were defined as the followings: initial stage,medium stage,and vigorous stage. The cluster analysis showed that when the proportions of the three phases of the initial,medium,and vigorous periods reached 13. 85%,29. 88%,and 56. 27%,respectively,the gas utilization rate was the largest,and the proportion of time during the vigorous gas flow was negatively correlated with the gas utilization rate. Further research shows that the gas flow center point was generated from the fourth circle elliptical from the southwest position of the physical center of the blast furnace,and its changing trajectory is a reciprocating motion between the southwest position and the physical center. The method proposed herein can help understand the dynamic changes of gas flow distribution in the blast furnace production process in real time,and then realize real-time monitoring and regulation of gas utilization rate.
引文
[1]宋建成.高炉炼铁理论与操作[M].北京:冶金工业出版社,2005,1.
[2]涂春林,毕学工,周勇.高炉炉顶温度分布模式识别神经元网络的研究[J].河南冶金,2004,12(1):10-13.
[3]薛崇盛,曹卫华,吴敏,等.高炉料面煤气流分布识别方法[J].清华大学学报(自然科学版),2008,48(10):1785-1789.
[4]徐化冰,徐华岩.基于TD算法的时序神经网络高炉炉喉十字测温预报[J].工业炉,2013,35(03):44-48.
[5]唐振浩,唐立新,杨阳.基于数据驱动和智能优化的高炉十字测温温度预报[J].信息与控制,2014,43(03):355-360.
[6]范志刚,邱贵宝,贾娟鱼,等.基于BP神经网络的高炉焦比预测方法[J].重庆大学学报(自然科学版),2002,(06):85-87,+91.
[7]王玉涛,姜慧研,徐小慧,等.混合神经网络及其在高炉径向煤气流模式分布中的应用[J].沈阳工业大学学报,1999,(3):265-268.
[8]崔桂梅,曹安威,马祥.一种基于感红外图像处理的高炉中心煤气流分布模式识别方法[J].信息与控制,2014,43(01):110-115.
[9]石琳,刘文磊,李江鹏,等.基于高炉炉喉煤气分布的煤气利用率预测模型[J].内蒙古科技大学学报,2016,(2):121-125.
[10] Shi L,Wen Y B,Zhao G S,et al. Recognition of Blast Furnace Gas Flow Center Distribution Based on Infrared Image Processing[J]. Journal of Iron and Steel Research(International),2016,23(03):203-209.
[11]许永华,吴敏,曹卫华,等.高炉温度场的红外图像识别检测方法及应用[J].控制工程,2005,(04):354-356.
[12]王昌军.高炉料面区域温度特征智能提取方法研究与应用[D].长沙:中南大学,2010.
[13]张重生.大数据分析:数据挖掘必备算法示例详解[M].北京:机械工业出版社,2016,12.
[14]周志华.机器学习[M].北京:清华大学出版社,2016,1.
[15]杨常清,王孝通.雷达图像像素矩阵的提取与应用仿真[J].计算机仿真,2003,(8):88-89.
[16]朱华东,宋威.基于像素差分基元矩阵的图像检索[J].计算机应用研究,2015,(10):3151-3155.
[17]黄敏,刘谦.基于矩阵像素阈值判定的运动目标检测[J].郑州轻工业学院学报(自然科学版),2010,(6):87-90.
[18]吴敏,王昌军,安剑奇,等.基于料面温度场的高炉煤气流分布识别方法[J].信息与控制,2011,40(01):78-82.
[19]杨丹,赵海滨,龙哲,等. MATLAB图像处理实例详解[M].北京:清华大学出版社,2013,7.
[20]江明,刘辉,黄欢.图像二值化技术的研究[J].软件导刊,2009,(04):175-177.
[2]涂春林,毕学工,周勇.高炉炉顶温度分布模式识别神经元网络的研究[J].河南冶金,2004,12(1):10-13.
[3]薛崇盛,曹卫华,吴敏,等.高炉料面煤气流分布识别方法[J].清华大学学报(自然科学版),2008,48(10):1785-1789.
[4]徐化冰,徐华岩.基于TD算法的时序神经网络高炉炉喉十字测温预报[J].工业炉,2013,35(03):44-48.
[5]唐振浩,唐立新,杨阳.基于数据驱动和智能优化的高炉十字测温温度预报[J].信息与控制,2014,43(03):355-360.
[6]范志刚,邱贵宝,贾娟鱼,等.基于BP神经网络的高炉焦比预测方法[J].重庆大学学报(自然科学版),2002,(06):85-87,+91.
[7]王玉涛,姜慧研,徐小慧,等.混合神经网络及其在高炉径向煤气流模式分布中的应用[J].沈阳工业大学学报,1999,(3):265-268.
[8]崔桂梅,曹安威,马祥.一种基于感红外图像处理的高炉中心煤气流分布模式识别方法[J].信息与控制,2014,43(01):110-115.
[9]石琳,刘文磊,李江鹏,等.基于高炉炉喉煤气分布的煤气利用率预测模型[J].内蒙古科技大学学报,2016,(2):121-125.
[10] Shi L,Wen Y B,Zhao G S,et al. Recognition of Blast Furnace Gas Flow Center Distribution Based on Infrared Image Processing[J]. Journal of Iron and Steel Research(International),2016,23(03):203-209.
[11]许永华,吴敏,曹卫华,等.高炉温度场的红外图像识别检测方法及应用[J].控制工程,2005,(04):354-356.
[12]王昌军.高炉料面区域温度特征智能提取方法研究与应用[D].长沙:中南大学,2010.
[13]张重生.大数据分析:数据挖掘必备算法示例详解[M].北京:机械工业出版社,2016,12.
[14]周志华.机器学习[M].北京:清华大学出版社,2016,1.
[15]杨常清,王孝通.雷达图像像素矩阵的提取与应用仿真[J].计算机仿真,2003,(8):88-89.
[16]朱华东,宋威.基于像素差分基元矩阵的图像检索[J].计算机应用研究,2015,(10):3151-3155.
[17]黄敏,刘谦.基于矩阵像素阈值判定的运动目标检测[J].郑州轻工业学院学报(自然科学版),2010,(6):87-90.
[18]吴敏,王昌军,安剑奇,等.基于料面温度场的高炉煤气流分布识别方法[J].信息与控制,2011,40(01):78-82.
[19]杨丹,赵海滨,龙哲,等. MATLAB图像处理实例详解[M].北京:清华大学出版社,2013,7.
[20]江明,刘辉,黄欢.图像二值化技术的研究[J].软件导刊,2009,(04):175-177.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |