基于高斯过程构建钢轨轧制力模型的方法
|
摘要
在钢轨轧制过程中,轧制力一定程度上对轧制参数调整有着指导或参考作用,一个成材率高、缺陷率低的轧制过程,必然会对应一个合理的轧制力分布。轧制力上冲过大不仅会使轧制状态靠近轧制系统的极点,稳定性降低,而且会导致轧辊磨损、疲劳加快,并将导致钢轨表面质量、规格波动等问题。采用高斯过程回归算法,基于现场生产数据建立了压下量与轧制力模型,为钢轨轧制过程轧制力的精确控制,提供了一种残差较小的统计模型。
In the process of rail rolling,the rolling force can guide or adjust the rolling parameters to a certain extent.A rolling process with high yield and low defect rate is bound to correspond to a reasonable distribution of rolling force.Too large rolling force will not only make the rolling state near the pole of the rolling system,but also decrease the stability,and will cause roll wear and fatigue to accelerate in a long time.The problems will be followed,such as rail surface quality,specification fluctuation and so on.The Gauss process regression algorithm was used in this paper,based on the field production data,the model of the reduction and rolling force was set up,which provides a statistical model for the precise control of the rolling force in the rail rolling process.
In the process of rail rolling,the rolling force can guide or adjust the rolling parameters to a certain extent.A rolling process with high yield and low defect rate is bound to correspond to a reasonable distribution of rolling force.Too large rolling force will not only make the rolling state near the pole of the rolling system,but also decrease the stability,and will cause roll wear and fatigue to accelerate in a long time.The problems will be followed,such as rail surface quality,specification fluctuation and so on.The Gauss process regression algorithm was used in this paper,based on the field production data,the model of the reduction and rolling force was set up,which provides a statistical model for the precise control of the rolling force in the rail rolling process.
引文
[1] 郑君里,应启珩,杨为理.信号与系统[M].北京:高等教育出版社,2011.
[2] 吴大正.信号与线性系统分析[M].北京:高等教育出版社,2005.
[3] 潘仲明.随机信号与系统[M].北京:国防工业出版社,2013.
[4] 王学颜,宋广惠.结构疲劳强度设计与失效分析[M].北京:兵器工业出版社,1992.
[5] 喜荣,陈林.轧钢工艺学[M].北京:冶金工业出版社,2013.
[6] 周建男.轧钢机[M].北京:冶金工业出版社,2009.
[7] 申倩情,孙宗海.基于自适应自然梯度法的在线高斯过程建模[J].计算机应用研究,2011,28(1):95.
[8] 刘开云,刘保国,徐冲.基于遗传-组合核函数启斯过程回归算法的边坡非线性变形时序分析智能模型[J].岩石力学与工程学报,2009,28(10):2128.
[9] 苏国韶,张研,燕柳斌.隧道围岩变形预报的高斯过程机器学习模型[J].桂林理工大学学报,2010,30(4):551.
[10] 何志昆,刘光斌,赵曦晶,等.高斯过程回归方法综述[J].控制与决策,2013,28(8):1121.
[2] 吴大正.信号与线性系统分析[M].北京:高等教育出版社,2005.
[3] 潘仲明.随机信号与系统[M].北京:国防工业出版社,2013.
[4] 王学颜,宋广惠.结构疲劳强度设计与失效分析[M].北京:兵器工业出版社,1992.
[5] 喜荣,陈林.轧钢工艺学[M].北京:冶金工业出版社,2013.
[6] 周建男.轧钢机[M].北京:冶金工业出版社,2009.
[7] 申倩情,孙宗海.基于自适应自然梯度法的在线高斯过程建模[J].计算机应用研究,2011,28(1):95.
[8] 刘开云,刘保国,徐冲.基于遗传-组合核函数启斯过程回归算法的边坡非线性变形时序分析智能模型[J].岩石力学与工程学报,2009,28(10):2128.
[9] 苏国韶,张研,燕柳斌.隧道围岩变形预报的高斯过程机器学习模型[J].桂林理工大学学报,2010,30(4):551.
[10] 何志昆,刘光斌,赵曦晶,等.高斯过程回归方法综述[J].控制与决策,2013,28(8):1121.