转炉冶炼终点锰成分的预报模型

[Mn]ep Prediction Model for Melt in Oxygen Converter

研究了转炉冶炼终点锰成分的影响因素,确定了预报模型的控制变量,建立了基于神经网络和自适应模糊神经网络的两种终点锰成分的预报模型,并对其进行了比较。研究发现,基于自适应模糊神经网络的预报模型能够很好地实现对终点锰成分的预报,在w([Mn])偏差值为±0.025％的控制精度范围内,预报命中率达到85.29％;在w([Mn])的偏差率为±25％范围内,预报的终点命中率达到70.59％。该模型接近基于副枪的终点锰成分动态预报模型的控制水平。

The influence of different steelmaking factors on [ %Mn]EP was studied and control variables for the prediction model determined and thereby two different models, one neural net based prediction model and one adaptive fuzzy-neural net based prediction model have been simultaneously established and compared with each other. Results show that the latter is much superior over the former in [Mn]EP prediction for the oxygen converter process and the hit rates of this model in ±0.025 % [Mn]EP deviation and in ±25 % [Mn]EP deviation rate are 85.29 % and 70.59 % respectively. The control performance of the model is close to that of the dynamic model for the sub-lance.