论文部分内容阅读
钢液成分复杂,含有多种元素,如何更好地检测出钢铁在冶炼过程中元素含量的变化,一直是冶金行业十分关注的问题。实时高效地检测钢液元素的含量变化有利于控制冶炼的终点,有利于节能减排。非线性回归分析用于分析因变量与自变量之间的非线性关系,钢液组成复杂,基体效应严重,非线性明显,本文利用多元非线性回归模型结合LIBS技术对钢液中的Mn元素进行定量分析。分析结果表明,验证集相对误差分别为7.273%,9.489%。因此,模型可用于对钢液中Mn元素的定量分析,预测精度较高,适用性较强。
The composition of molten steel is complex, contains a variety of elements, how to better detect changes in the elemental content of steel in the smelting process, has been the metallurgy industry is very concerned about the problem. Real-time and efficient detection of changes in the content of molten steel is conducive to controlling the end of smelting, is conducive to energy saving. Nonlinear regression analysis is used to analyze the nonlinear relationship between dependent variable and independent variable. The composition of molten steel is complex and the matrix effect is serious with obvious nonlinearity. In this paper, Multivariate nonlinear regression model combined with LIBS Quantitative analysis. The analysis results show that the relative errors of the verification set are 7.273% and 9.489% respectively. Therefore, the model can be used for quantitative analysis of Mn in molten steel with high prediction accuracy and strong applicability.