神经网络在高炉中的应用  

作　　者：张涛[1] 李福进[2] 

机构地区：[1]河北理工大学 [2]中国矿业大学(北京校区)

出　　处：《信息通信》2007年第1期46-49,共4页Information & Communications

摘　　要：文章介绍了神经网络在高炉中的应用,将导致高炉操作炉型异常及炉温分布异常的征兆作为输入量,分别采用不同的子神经网络,完成不同侧面高炉异常运行模式的识别。用决策级融合网络得出综合决策结论,能够充分利用不同信息源的信息,提高了决策结论的确定性,从而获得对高炉运行状态的最优估计与判决。在理论研究和仿真验证基础上,基于MATLAB及VB软件技术建立了高炉智能操作决策系统的工程应用方法,为高炉智能操作决策系统的在线应用作了技术上的准备。The article introduced the application of neural network in the blast furnace, taking the cause of the blast furnace exceptional operation and the fumace temperature unusual distribution indication as the input values, and by means of a sub- neural network separately it realized the recognition of operation pattern on different sides of the blast furnace. Comprehensive decision- making conclusion drawn by means of policy-making level integrated network can make full use of information from different sources. The emulational analysis indicates that the integrated decision-making system can reduce the uncertainty of decision-making conclu- sion thus the most accurate estimate and the judgment of the operating state of the blast furnace can be guaranteed. It has established the project application method to blast furnace intelligence operation decision system based on MATLAB and the VB technology, which provides preparation for the blast fumace intelligence operation decision system to be served on line.

关 键 词：高炉 神经网络 智能决策 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]