摘要
利用神经网络可以以任意精度逼近任意复杂非线性函数的特点,建立了铁矿粉烧结基础特性的预报模型,并对常用铁矿粉中的10种矿的同化性、液相流动性和粘结相强度进行了预报,模型的准确度达到了75%~90%。该预报模型的构建对于烧结配矿决策、提高烧结矿质量具有指导意义。
Using the characteristics that neural networks can approximate any arbitrary precision with any complex nonlinear function,a model was established to predict the iron ore sintering basic characteristics,and then forecast the assimilation,the flow of liquid and the strength of the binding phase,the accuracy achieved 75% ~ 90% . This prediction model has guiding significance for sinter ore decision and improving the quality of sinter.
出处
《河北联合大学学报(自然科学版)》
CAS
2014年第3期16-18,共3页
Journal of Hebei Polytechnic University:Social Science Edition
关键词
铁矿粉
化学成分
烧结基础特性
神经网络
iron ore powder
chemical composition
the sintering basic characteristics
neural network