无缝钢管张力减径是多机架机组对荒管进行无芯棒连轧过程。它具有复杂的金属流动状态,影响其精度的工艺参数较多,难以用现有轧制理论进行分析。应用人工神经网络的BP算法,对不同坯料样本进行训练学习,确定其权值和阈值,建立起预测管材精度的数学模型。经试验证明该模型具有较高的可靠性和精度。 Seamless steel pipe tension reduction is a multi-machine crew of non-core rod waste bar continuous rolling process. It has a complex metal flow state, affecting the accuracy of the process parameters are more difficult to analyze with the existing rolling theory. The BP algorithm of artificial neural network is used to train and learn the samples of different billets. The weights and thresholds of the billets are determined, and a mathematical model for predicting the precision of the pipe is established. The experiment shows that the model has high reliability and precision.