摘要
利用某钢厂高炉热风炉历史数据,提出一种基于神经网络遗传算法寻求高炉热风炉最佳空燃比的方法。文中结合热风炉操作工艺及专家经验分析了高炉热风炉空燃比、拱顶温度与热效率的关系,并通过BP神经网络建立了准确的描述拱顶温度与包括空燃比在内变量的非线性函数关系,最后利用遗传算法进行寻优计算得出最佳空燃比的参考值。仿真结果表明,该方法建立的非线性函数关系准确,通过遗传算法计算得到的空燃比能显著提高热风炉燃烧热效率。
Using the historical data of the blast furnace,a method was proposed to find the best air-fuel ratio based on neural network and genetic algorithm. This paper combine with the blast furnace operation process and expert experience to analyze the relationship between air-fuel ratio,dome temperature and combustion efficiency.The BP neural network was used to establish a nonlinear function.The function can express the relationship between the variables accurately.Finally,use genetic algorithm to optimize the best air-fuel ratio reference value.The simulation results show that the nonlinear function relationship was accurate,and the air-fuel ratio can significantly improve the thermal efficiency of blast furnace.
引文
[1]景会成,于玉超,侯宝稳,等.基于神经网络的加热炉烟气含氧量预测及控制[J].自动化与仪表,2012,27(2):30-32.
[2]孙进生.高炉热风炉燃烧CBR智能控制技术[M].北京:冶金工业出版社,2006.
[3]赵玉贵,张明金.基于数学模型的出钢温度预报数值模拟[J].自动化与仪表,2013,28(3):10-12.
[4]王士超,白瑞祥.基于PLC的锅炉燃烧模糊自寻优控制[J].自动化与仪表,2008,23(1):33-36.
[5]湛腾西,张国云,胡文静.基于模糊满意度的热风炉空燃比优化方法[J].冶金自动化,2008,32(6):1-6.
[6]马竹梧.高炉热风炉全自动控制专家系统[J].控制工程,2002,9(4):57-62.
[7]Yunosuke Maki,Akihiro Inayama,Katsumi Ino.The lastest technologies for process control and automation in blast furnace[J].KAWASAKI Steel Technical Report,2000,1(43):61-67.
[8]秦伟,陶新娅.热风炉高风温操作影响因素分析[J].鞍钢技术,2009(2):35-38.