基于料面温度场特征的高炉炉况诊断方法

Blast furnace status diagnosis based on burden surface temperature field feature

针对高炉(BF)生产过程中炉况状态难以诊断的问题,提出一种基于料面温度场特征信息的高炉炉况最小二乘支持向量机(LS-SVM)诊断方法.根据高炉炉顶红外图像和过程检测数据,采用信息融合技术建立高炉料面温度场.提取与炉况有关的料面温度场的特征信息,将特征信息作为炉况诊断模型的输入.采用最小二乘支持向量机构造炉况诊断模型,计算高炉运行状态.现场数据验证表明,该方法充分利用了温度场和炉况的关系,具有较好的诊断效果,能够有效地帮助现场操作人员及时发现异常炉况,稳定高炉运行.

In order to resolve the diagnosis difficulty of blast furnace（BF） status,a diagnosis method for BF status was proposed based on the least square support vector machine（LS-SVM） method and the feature of burden surface temperature field.The burden surface temperature field was conducted by using the information fusion technology according to the infrared image in BF top and the process variable.The input of BF status diagnosis model was achieved based on the feature information related with the BF status which was extracted from the burden surface temperature field.The LS-SVM was employed to construct a model to diagnose the BF status.The field,data show that the method fully uses the relationship between the BF status and the temperature field,and has a better diagnosis effect,which helps operators detect the abnormal BF status and stabilize the BF operation.