基于神经网络烧结终点(BTP)预测系统  

ADAPTIVE PREDICTION OF SINTER BURNING THROUGH POINT BASED ON NEURAL NETWORK

作　　者：程武山[1] 刘长青[2] 

机构地区：[1]上海大学机械工程和自动化学院,上海200072 [2]马鞍山钢铁公司第三烧结厂,马鞍山243005

出　　处：《计算机应用与软件》2003年第7期39-40,94,共3页Computer Applications and Software

摘　　要：为解决烧结过程大滞后环节和烧结终点难以测量的困难 ,采用MBP神经网络对烧结过程进行预估、仿真和测试 ,建立了有较好泛化能力的系统模型 ,实现了对烧结终点的提前预报和生产操作指导。In order to overcome the long time delay and Sinter Burining Through Point of Sinter Process,a better wide adaptive system model was established by using MBP neual network to predict,simulate and test sinter process.The proposed predictor can be used to predict future value of Burning Through Point(BTP) and to help operator to adjust sinter process.

关 键 词：烧结终点 预测系统 神经网络 学习算法 烧结机 日本川崎钢铁公司水岛厂 

分 类 号：TF046.4[冶金工程—冶金物理化学]