基于DBO-RBF的无缝钢管张力减径轧制力预测  

Prediction of tension reduction rolling force of seamless steel pipe based on DBO-RBF

作　　者：胡建华[1] 马佳旺 黄宇龙 郝亚栋 张根耀 裴艺航 HU Jian-hua;MA Jia-wang;HUANG Yu-long;HAO Ya-dong;ZHANG Gen-yao;PEI Yi-hang(School of Materials Science and Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学材料科学与工程学院,山西太原030024

出　　处：《塑性工程学报》2024年第11期90-96,共7页Journal of Plasticity Engineering

基　　金：山西省科技重大专项(20191102009)。

摘　　要：基于深度学习,提出了一种基于蜣螂优化算法(DBO)优化径向基函数(RBF)神经网络的预测模型,对无缝钢管张力减径过程的轧制力进行预测。同时,以某钢管厂张力减径过程中采集的第4和第10机架轧制力相关数据为样本集,通过学习和训练,将得到的结果与传统RBF神经网络的预测结果进行了对比。结果表明,该模型具有更高的预测精度和稳定性,第4机架轧制力预测结果的均方根误差eRMSE和平均绝对误差e_(MAE)分别为0.53和0.39,第10机架轧制力预测结果的均方根误差e_(RMSE)和平均绝对误差e_(MAE)分别为0.13和0.09,预测值与真实值之间误差均在工业允许范围内。Based on deep learning,a prediction model based on dung beetle optimization(DBO)algorithm to optimize radial basis function(RBF)neural network was proposes to predict the rolling force of seamless steel pipe during tension reduction process.At the same time,the data related to the rolling force of the fourth and the tenth stands collected in the tension reduction process of a steel pipe mill were taken as the sample set,and the results obtained by learning and training were compared with the prediction results of the traditional RBF neural network.The results show that the model has higher prediction accuracy and stability,the root mean square error e_(RMSE) and mean absolute error e_(MAE) of the fourth stand rolling force prediction results are 0.53 and 0.39,respectively,and the root mean square error e_(RMSE) and mean absolute error e_(MAE) of the tenth stand rolling force prediction results are 0.13 and 0.09,respectively,and the errors between the predicted values and the true values are within the allowable range of the industry.

关 键 词：蜣螂优化算法 径向基函数神经网络 无缝钢管 张力减径 轧制力预测 

分 类 号：TG335.71[金属学及工艺—金属压力加工]