基于k-NN模型的钢坯入炉温度预测  

Prediction of Billet Charging Temperature Based on k-NN Model

作　　者：张仁琳 Zhang Renlin(Plate Factory,Fujian Sanming Iron and Steel(Group)Co.,Ltd.,Sanming 365000,Fujian)

机构地区：[1]福建三钢(集团)有限责任公司中板厂,福建三明365000

出　　处：《福建冶金》2023年第3期24-26,共3页Fujian Metallurgy

摘　　要：在中厚板生产中,加热工序是决定后续轧制过程是否稳定的重要环节。同时为了提高经济效益,实现节能减排,加热工序往往会根据原料本身的温度情况来制定相应的加热工艺。换热系数是建立钢坯温降计算模型的核心参数,由于其影响因素繁多且复杂,很难有一个固定的模型来计算。本文提出一种简单有效的基于k-NN算法的温度预测模型,寻找目标钢坯与样本钢坯之间的相似度,然后通过IDW加权平均算法,预估出目标钢坯的入炉温度。实际应用表明,模型预测温度与实测温度绝对误差控制在30℃以内的占比达95%以上。In the production of medium and heavy plates,the heating process is an important link in determining whether the subsequent rolling process is stable.In order to improve economic benefits and achieve energy saving and emission reduction,the heating process often formulates corresponding heating processes according to the temperature of the raw materials themselves.The heat transfer coefficient is the core parameter for establishing the billet temperature drop calculation model.Due to its many and complex influencing factors,it is difficult to have a fixed model to calculate.In this paper,a simple and effective temperature prediction model based on the k-NN algorithm is proposed to find the similarity between the target billet and the sample billet,and then use the IDW weighted average algorithm to estimate the temperature of the target billet.The practical application shows that the absolute error between the model predicted temperature and the measured temperature is controlled within 30℃,accounting for more than 95%.

关 键 词：钢坯温度 k-NN模型 温度预测 

分 类 号：TG307[金属学及工艺—金属压力加工]