高炉料面区域温度特征智能提取方法研究与应用

【作者】 王昌军；

【导师】 吴敏； 何勇；

【作者基本信息】 中南大学 ， 控制科学与工程， 2010， 硕士

【摘要】 针对高炉内部环境复杂,反映煤气流发展状态的料面温度特征难以提取的问题,本文从工艺机理的角度分析了料面红外图像、十字测温、上升管温度、炉墙温度等与料面温度场之间的关系,根据煤气流分布特点,提出了一种高炉料面区域温度特征提取方法,分别对高炉中心区域和边缘区域的特征进行提取。首先,进行料面中心区域温度特征提取。根据红外摄像仪的成像原理与安装情况,对红外图像和高炉料面的坐标系完成空间配准；依据高炉生产特点,对高炉中心点进行离线配准,采用动态定标算法,完成红外可视区的温度场检测。利用图像阈值分割与空间配准,进行炉心区宽度和炉心偏移程度特征进行提取。结合料面中心温度指数与炉心宽度,采用模糊推理的方法对中心区的温度特征进行评价,反映中心煤气流的发展状态。其次,进行料面边缘区域的温度特征提取。由于十字测温、炉墙温度以及矿焦比与料面边缘温度存在相关性,首先采用单一信息在不同尺度下对边缘温度进行检测；然后,采用模糊D-S证据方法融合单一信息的检测值,获得料面边缘温度,降低了检测的不确定性；最后,结合料面边缘温度指数与边缘宽度,采用模糊推理的方法对边缘区的温度特征进行评价,反映边缘煤气流的发展状态。根据提出的特征提取方法,开发了高炉料面温度场检测系统,并成功运行于某钢铁企业2200m3高炉。系统采用可视化界面显示料面温度场,更直观地反映料面温度分布情况,为高炉操作人员提供了炉况的实时、可靠的参考信息。更多还原

【Abstract】 With the problems about complex environment inside blast furnace(BF) and the difficulty in extracting temperature features of burden surface which reflect the status of gas flow, the relationships among temperature fields of burden surface, infrared image of burden surface, crossing temperature, temperature of rising pipes, and temperature of wall are analyzed in this thesis. According to the characteristic of gas flow distribution, a method is proposed to extract the temperature features in central and peripheral regions of burden surface.Firstly, the temperature features in the central region of burden surface are extracted. The coordinates between infrared image and burden surface are matched according to imaging principle of camera and installation conditions of infrared camera, and central point of BF is matched offline in accordance with production process. Then the way of dynamic temperature calibration is employed to achieve the temperature fields of infrared visible region. Through the methods of image threshold segmentation and spatial registration, the features about width and deviation extend of central region are extracted. With temperature index of center and width of central region, the method of fuzzy reasoning is employed to evaluate the feature of central temperature which displays the developing status of central gas flow.Secondly, the temperature features in the peripheral region of burden surface are extracted. Beacause of the correlations between temperature of peripheral region and some information such as crossing temperature, wall temperature and ratio of ore and coke, the single information is utilized to measure the temperature of peripheral region of burden surface with different scales. Then the method of fuzzy D-S evidence is adopted to fuse the measured results from singe information, which may reduce the uncertainty. With the peripheral temperature index and width of peripheral region, fuzzy reasoning is also applied to evaluate the feature of peripheral temperature which displays the developing status of peripheral gas flow.Finally, according to the method of features extraction proposed in this thesis, a real-time monitoring system of burden surface temperature fields in BF is developed and successfully applied in a 2200m3 BF in some steel company. The visual interface of monitoring system is more effective to understand the distribution of burden surface temperature field and guide the operation of burden distribution.更多还原