基于神经网络的连铸钢包下渣识别系统  

Identification System for Slag Detection of Continuous Casting Based on Neural Network

作　　者：陈至坤[1] 魏殿才[1,2] 王福斌[1] 

机构地区：[1]河北联合大学电气工程学院,河北唐山063009 [2]唐山劳动技师学院电气工程系,河北唐山063009

出　　处：《机械工程与自动化》2011年第5期120-122,共3页Mechanical Engineering & Automation

摘　　要：BP网络可以逼近复杂的非线性函数,可广泛用于函数拟合、模式识别、分类、数据压缩、预测等。设计了连铸钢包下渣自动检测系统中BP神经网络的结构,并结合钢包下渣工艺参数,对钢包下渣时刻进行了预报分析。仿真结果证明BP神经网络能够提高连铸钢包下渣时刻预报的准确性。BP network can approach the complex nonlinear function,and is widely used in function fitting,pattern recognition,classification,data compression and forecasting,etc.In this paper,the BP network structure for slag automatic detection of continuous casting is designed.Analyses of the ladle slag moment are carried out by taking account of slag ladle technological parameters,The simulation results prove that BP neural network can improve the accuracy of forecasting of ladle slag moment.

关 键 词：连铸下渣 BP神经网络 识别系统 

分 类 号：TF777[冶金工程—钢铁冶金] TP183[自动化与计算机技术—控制理论与控制工程]