首页|一种液驱冷轧机辊热凸度动态SVR预测模型

一种液驱冷轧机辊热凸度动态SVR预测模型

扫码查看
辊热凸度是当轧件发生变形或因摩擦作用生成热量而引起轧辊温度上升并由此产生轧辊外形尺寸改变的现象.为了降低液驱冷带轧机辊随温度的变化影响,综合分析了辊热凸度因素引起的变形效果差异性.选择液驱轧机作为测试对象,从热凸度作用原理方面考虑,同时根据实际获得的监测参数探讨了辊热凸度情况受到外部因素的影响程度.建立基于支持向量机的液驱冷带轧机辊热凸度动态预测模型,该模型结构较简单,并且可以获得优异的性能.对于辊热凸度的预测表现出了明显的优势.利用SVR模型以及交叉验证的方式确定参数,结果形成了一致的变化趋势与周期性特征.拟合优度与均方根误差分别为0.84与0.0153,获得了较理想的预测性能.该研究对提高板材的控制精度以及轧机节能具有较好的实际指导意义.
A Dynamic SVR Prediction Model for Thermal Crown of Liquid Driven Cold Rolling Mills
The thermal crown of the roll is the phenomenon that the shape of the roll changes due to the rise of the temperature of the roll caused by the deformation of the rolled piece or the heat generated by the friction.In order to reduce the influence of the change of temperature on the rolls of the hydraulically driven cold strip mill,the deformation effect difference caused by the ther-mal crown of the rolls was analyzed comprehensively.The hydraulic drive mill was chosen as the test object,and the influence of external factors on the thermal crown of the roll was discussed based on the principle of thermal crown and actual monitoring pa-rameters.A dynamic prediction model of thermal crown of hydraulic drive cold strip mill based on support vector machine is estab-lished.The prediction of roller thermal crown has obvious advantages.The parameters were determined by SVR model and cross-validation,and the results formed consistent trend and periodicity.The goodness of fit and root mean square error are 0.84 and 0.0153 respectively,and the better prediction performance is obtained.The research has a good practical guiding significance for improving the control precision of plate and saving energy of rolling mill.

Cold Strip MillRoller Thermal CrownSupport Vector MachineDynamic Prediction

车智峰、张小菊、王国强

展开 >

新乡职业技术学院电子信息系,河南 新乡 453000

河南理工大学计算机科学与技术学院,河南 郑州 450000

河南英科传感技术有限公司,河南 郑州 450000

冷带轧机 辊热凸度 支持向量机 动态预测

河南省高等学校重点科研项目

21B460003

2024

机械设计与制造
辽宁省机械研究院

机械设计与制造

CSTPCD北大核心
影响因子:0.511
ISSN:1001-3997
年,卷(期):2024.398(4)
  • 13