基于模式识别的自学习型高炉冶炼专家系统的开发与应用  

Development and application of blast furnace expert system with self-learning function based on pattern recognition

作　　者：陈令坤[1] 李佳[2] 

机构地区：[1]武汉钢铁集团公司研究院,武汉430081 [2]武汉科技大学机械自动化学院,武汉430081

出　　处：《东南大学学报（自然科学版）》2012年第A01期117-121,共5页Journal of Southeast University：Natural Science Edition

摘　　要：为了准确反映高炉冶炼过程的复杂性,实现高炉过程的准确控制,基于模式识别的理论设计了一种高炉冶炼专家系统,利用该系统可以实现对关键控制参数如风量、风压、风温等进行模式识别,在此基础上能够实现高炉过程的总体评估,同时利用模式识别技术可以对影响高炉过程的重要现象,如炉顶煤气流分布、炉型变化、布料等进行有效的分类管理,实现对高炉过程的准确控制,从而提高了高炉运行的稳定性,高炉的技术经济指标有明显改善,2010年8月高炉专家系统在武钢5号高炉投运后,降低焦比8.42 kg/t,节约焦炭7732.3 t;增产34142 t.In order to reflect the complexity of blast furnace and achieve precision control of blast furnace, a blast furnace expert system based on pattern recognition is developed. The characteristics of key parameters, such as the blast volume, the blast pressure and the blast temperature, can be classified based on pattern recognition, and the total condition of blast furnace can be evaluated. Some phenomenon, such as the top gas distribution, the variation of inner profile of blast furnace and burden distribution, are classified based on pattern recognition and the precision control of the blast furnace can be realized. So the stability of blast furnace can be achieved by using the blast fur nace expert system. This system has been used in blast furnace No. 5 in Wuhan Iron ＆ Steel（Group） Corporation since August 2010 with a decreased coke rate of 8.42 kg/t, the coke of 7732.3 t has been saved, and the amount of increase in production of hot meatal is about 34 142 t in half a year.

关 键 词：高炉 模式识别 布料 气流控制 炉型管理 

分 类 号：TQ542.7[化学工程—煤化学工程]