基于TD算法的时序神经网络高炉炉喉十字测温预报  

Blast Furnace Throat Cross Temperature Forecast by Time-Sequence Neural Network Based on TD Algorithm

作　　者：徐化冰[1] 徐华岩[2] 

机构地区：[1]渤海船舶职业学院动力工程系,辽宁葫芦岛125000 [2]冶金自动化研究设计院,北京100071

出　　处：《工业炉》2013年第3期44-48,共5页Industrial Furnace

摘　　要：通过对高炉煤气流的分析,给出了影响高炉炉喉温度变化的状态量;采用基于TD算法的时序神经网络进行十字测温的预报,并通过模拟现场环境进行实验。结果表明,利用TD算法和人工神经网络来预测高炉炉喉十字测温基本可行,预测误差较小,精度较高。The state-arguments forward through analyzing the blast influencing the blast furnace furnace gas flow. Using the throat temperature change were put time-sequence neural network cross temperature prediction based on the TD algorithm, the experiment was carried out through simulating the on-site environment. The results show that it is basic feasible to predict the blast furnace throat cross temperature by using the TD algorithm and artificial neural network. The prediction error is small and the precision is high.

关 键 词：炼铁高炉 TD算法 神经网络 十字测温 

分 类 号：TF54[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]