论文部分内容阅读
本文针对硫的优化问题建立了神经网络模型。影响铁水中的硫含量因素有很多,主要从渣氧势含量、炉渣碱度方面进行了分析,利用时差法,对铁水中硫含量进行预测与优化:渣氧势越低,对炉渣除硫的反应越有利,硫含量越低;在一定限度内,Cao、Mgo含量越高,TiO2含量越低,对复杂碱度的增加量影响越大,硫含量越低,研究提高了其命中率,在此基础上利用问题一的模型对硅含量进行了合理的期望预测,精度较高。
In this paper, a neural network model is established for the optimization of sulfur. There are many factors affecting the sulfur content in molten iron, mainly from the slag oxygen potential, slag alkalinity aspects were analyzed using the time difference method to predict and optimize the sulfur content in molten iron: the lower the slag oxygen potential of the slag sulfur removal The more favorable the reaction, the lower the sulfur content; within a certain limit, the higher the content of Cao, Mgo, the lower the content of TiO2, the greater the impact on the increase of complex alkalinity, the lower the sulfur content, Based on the model of Problem 1, this paper makes a reasonable prediction of the silicon content with high precision.