摘要
文章针对BP网络收敛速度慢和易于陷入极小值的问题,应用RBF网络模型对高炉铁水硅含量进行了预测,通过对高炉一段连续时期内正常生产的数据经过归一化处理后进行训练和仿真,结果表明,高炉冶炼在运用了先进的RBF人工神经网络预测模型后,能预测铁水硅含量的高低,从而判断炉温走势,调控炉温,同时监测多个生产过程控制对象,有利于提高高炉生产艺,实现节能降耗。
The model of Radial Basis Function (RBF) neutral network is used to predict the silicon content of blast furnace hot metal as the calculation accuracy and convergence rate of the BP neural network are limited. Retreated data of the blast furnace is used to train and simulate the model. The simulation results are proved that the model of RBF can forecast Si-content so as to judge the trend of furnace temperature and control furnace temperature. What's more, the model can monitor multi-objects in production process at the same time. It is helpful to improve production process of blast furnace and reduce the energy consumption.
出处
《微计算机信息》
2009年第14期232-233,258,共3页
Control & Automation
基金
基金申请人:邱东
计划名称:国家科技支撑计划(2007BAQ00097)
基金申请人:邱东项目名称:冶金企业氩氧精炼铁合金工艺及综合节能技术的开发与示范(2007BAQ00097-4)
基金申请人:邱东项目名称:冶炼过程能耗监测与能耗系统优化
关键词
铁水硅含量
RBF神经网络
预测模型
能耗
hot metal Si-content
RBF neural network
prediction model
energy consumption