作　　者： ; ; ;

机构地区： 东北大学材料与冶金学院

出　　处： 《炼钢》 2006年第1期40-44,共5页

摘　　要： 准确预报转炉冶炼终点的钢水温度与碳含量对提高转炉终点命中率具有重要意义。针对现有多层前馈网络学习算法的不足,基于BP模型提出一种改进算法,建立了复吹转炉冶炼终点的预报模型,并与BP模型的预测结果进行了统计比较。研究表明,改进后的模型能够对冶炼终点进行良好的预报。采用单节点输出模型对终点钢水碳含量与温度分别进行预报,预测误差w(Δ[C])<±0.03%的命中率达97.22%,Δt<±12℃的命中率为94.44%。还建立了神经网络双节点输出模型对转炉终点钢水碳含量及温度同时进行预报,误差Δt<±15℃、w(Δ[C])<0.03%的双命中率为76.92%。 Accurate prediction of the end-point temperature and carbon content of BOF molten steel is of great significance to raising the hitting rate of the end-point. In view of the deficiencies of the present back-propagation algorithm a BP-based improved algorithm is put forward and then a prediction model for the end-point of the combinedblown converter refining is established and the prediction results of the new model are compared with that of the original one. Results show that the improved model can more effectively predict the BOF blow end-point. The carbon content and temperature of BOF blow end-point liquid steel have been predicted respectively using a single node output model. The hit rate of the model with w （△[C]）〈±0.03 %is about 97. 22 %and the hit rate of the model with △〈± 12℃ is about 94. 44 %. A double output neural network based model is established and the carbon content and temperature of the BOF blow end-point liquid steel are simultaneously predicted by the second model. The hit rate of the second model with w（△[C]）〈± 0. 03 % and △t〈±15℃ is about 76. 92 %.

关 键 词： 氧气转炉 神经网络 终点预报 碳含量 温度

领　　域： [冶金工程]