一种基于GRNN的LF炉温度预报模型  

A GRNN-based Temperature Prediction Model for LF Furnaces

作　　者：徐湛博 齐文哲 徐少川[1] Xu Zhanbo;Qi Wenzhe;Xu Shaochuan(School of Electronic and Information Engineering,University of Science and Technology Liaoning,Anshan,China)

机构地区：[1]辽宁科技大学电子与信息工程学院,辽宁鞍山

出　　处：《科学技术创新》2023年第8期64-67,共4页Scientific and Technological Innovation

摘　　要：钢包炉(LF)精炼是进行炉外精炼的主要手段。其主要任务之一就是对钢液的温度进行精确的控制与调整,因此需要及时获取钢液温度信息。本研究对某钢厂120吨精炼炉现有生产方式进行研究,对影响温度的因素进行分析,并对采集到的数据进行预处理。基于传统神经网络在小样本集上拟合能力差的问题,本研究选用广义回归神经网络建立模型,可以在提升预测实时性的同时保证精度。在与其他网络模型对比后证明广义回归神经网络的预测性能较好,准确度高。Ladle furnace(LF)refining is the main means of refining outside the furnace.One of its main tasks is the precise control and adjustment of the steel temperature,so it is necessary to obtain timely information on the steel temperature.In this research,the existing production method of a 120-ton refining furnace in a steel plant is studied,the factors affecting the temperature are analyzed,and the collected data are pre-processed.Based on the problem of poor fitting ability of traditional neural networks on small sample sets,this research selects generalized regression neural network to build the model,which can improve the prediction real-time while ensuring the accuracy.The prediction performance of generalized regression neural network is proved to be better and accurate after comparing with other network models.

关 键 词：锅包炉 钢液温度 广义回归神经网络 预测性能 

分 类 号：TF769[冶金工程—钢铁冶金]