无缝钢管斜轧穿孔管形多目标预测  

Multi-objective prediction on cross-rolling and perforated pipe shape for seamless steel pipe

作　　者：加世滢 王清华[1] 王贞艳[1] 胡建华[2] 双远华[2] 周新亮[3] Jia Shiying;Wang Qinghua;Wang Zhenyan;Hu Jianhua;Shuang Yuanhua;Zhou Xinliang(School of Electronic and Information Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China;School of Materials Science and Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China;Technical Center of Taiyuan Heavy Industry Co.,Ltd.,Taiyuan 030024,China)

机构地区：[1]太原科技大学电子信息工程学院,山西太原030024 [2]太原科技大学材料科学与工程学院,山西太原030024 [3]太原重工股份有限公司技术中心,山西太原030024

出　　处：《锻压技术》2022年第10期169-175,共7页Forging & Stamping Technology

基　　金：山西省科技重大专项(20191102009)。

摘　　要：针对无缝钢管斜轧穿孔生产中工艺参数对毛管尺寸精度的影响问题,考虑生产工艺以及生产需求优化等因素,建立了基于最小二乘支持向量回归的多目标预测模型。通过灰色关联分析法对影响因素进行分析筛选,以前伸量、轧辊间距、导板间距、顶头直径、坯料直径5个工艺参数作为预测模型的输入,以毛管壁厚和外径2个管形参数作为预测模型的输出;考虑到数据样本小且输入和输出参数之间的交叉相关性问题,构建了多输入多输出最小二乘支持向量回归模型对毛管管形进行预测。将实际采集的数据作为训练样本,通过仿真实验证明了模型的有效性,研究结果可为无缝钢管斜轧穿孔生产过程中所需的工艺参数调整与优化提供参考。In view of the influence of process parameters on the dimensional accuracy of capillaky in the cross-rolling and preforating production of seamless steel pipe,a multi-objective prediction model based on least squares support vector machine was established considering the factors of production process and production demand optimization,and the influencing factors were analyzed and screened by the grey correlation analysis method.Then,taking five process parameters of forward extension amount,roll spacing,guide plate spacing,plug diameter and blank diameter as inputs of the prediction model and the two pipe shape parameters of capillary wall thickness and outer diameter as outputs of the prediction model,considering the small data sample and the cross correlation between input and output parameters,a multi-input and multi-output least squares support vector regression model was constructed to predict the capillary shape.Furthermore,taking the actual data as training samples,the effectiveness of the model was proved by simulation experiments.And the research results provide reference for the adjustment and optimization of the process parameters required in the cross-rolling and preforating production process of seamless steel pipe.

关 键 词：无缝钢管 斜轧穿孔 最小二乘支持向量回归 管形预测 工艺参数 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]