摘要
针对传统带钢凸度预测方法预测精度低、速度慢的问题,建立了基于随机森林和支持向量机的热轧带钢凸度加权预测模型。采用改进长鼻浣熊算法分别对随机森林、支持向量机和随机森林与支持向量机加权预测模型的参数进行优化,提高凸度预测精度。以某公司热轧1 580 mm生产线实测数据进行凸度预测仿真研究,随机森林与支持向量机加权预测模型的均方根误差为2.23μm,与随机森林模型、支持向量机模型预测精度进行比较,加权预测模型的精度分别提高了7.08%、2.62%。
In view of low prediction accuracy and slow speed of traditional prediction methods for strip crown,a weighted prediction model based on random forest(RF) and support vector machine(SVM) was established.The parameters of models based on RF,SVM,and a combination of RF and SVM were optimized respectively by adopting the improved coati optimization algorithm(ICOA),so as to improve crown prediction accuracy.A 1 580 mm production line of a hot-rolling mill in one company was taken in a simulation research on crown prediction based on its actual measurement.The root mean square error of the weighted prediction model based on RF and SVM is 2.23 μm.It is found that this weighted prediction model has its prediction accuracy increased by 7.08% and 2.62% respectively,compared with the models based on RF and SVM respectively.
作者
周亚罗
李子轩
张少川
刘文广
张瑞成
ZHOU Yaluo;LI Zixuan;ZHANG Shaochuan;LIU Wenguang;ZHANG Ruicheng(College of Electrical Engineering,North China University of Science and Technology,Tangshan 063210,Hebei,China;Shougang Jingtang United Iron and Steel Co.,Ltd.,Tangshan 063200,Hebei,China)
出处
《矿冶工程》
CAS
北大核心
2024年第6期144-150,共7页
Mining and Metallurgical Engineering
基金
河北省自然科学基金资助项目(F2018209201)
唐山市科技局科技计划资助项目(22130213G)。
关键词
凸度预测
热轧带钢
支持向量机
长鼻浣熊算法
凸度
随机森林
crown prediction
hot rolling strip
support vector machine(SVM)
coati optimization algorithm(COA)
crown
random forest(RF)
