【摘要】 高炉铁水中的硫含量是描述铁水质量的一个重要指标.为了在出铁之前了解铁水中硫含量的高低,建立预测模型是必要的.本文利用遗传算法(GA)和BP神经网络构造了高炉铁水硫含量的预测分析模型,从某高炉选取117组数据进行学习和预测.运行结果表明,模型预测精度较高,当要求绝对误差为±3×10-6时,命中率可达61.54%;绝对误差为±4×10-6时,命中率可达84.69%.在此基础上,应用该模型回归分析了高炉风量、热风压力、富氧量、铁间料批数与铁水硫含量之间的相关关系,结果与高炉冶炼理论基本吻合,可为高炉生产提供一定的指导.更多还原

【Abstract】 Sulfur content is an important index to describe hot metal quality.It is necessary to build a model to predict sulfur content.In this paper,combining the Genetic Algorithms（GA） and Back-propagation neural network（BP）,an intelligent GA-BP model was established to predict the sulfur content in hot metal.117 datas were chosen to train the network model.The results showed that the model has fairly high accuracy.When the required absolute error was within ±3×10^-6,the accuracy of model can reach 61.54%;and when the absolute error was within ±4×10^-6,the accuracy can reach 84.69%.In addition,the relation between some operating parameters and sulfur content had been analysed,such as blast volume,blast pressure,charging batch, etc.The result was consistent with ironmaking theory.This can provide theoretical basis for ironmaking production.更多还原