In this paper, the research work of the torque model in plate rolling process, with biting impact considered, is carried out based on mechanical dynamics and the rolling process technology. The five-degree-of-freedom mechanical dynamic model was established for the main drive system of the actual heavy plate mill, considering the clearance between parts. The biting peak torques under different rolling process conditions were calculated. The influence of the biting time and the steady-state torque were analyzed: the biting peak torque decreases with the biting time increasing, and increases with the steady-state torque increasing. The biting time calculation model was established based on the rolling process parameters. The steady-state torque model was improved by rebuilding level arm ratio model. The influence of deformation area arithmetic average aspect ratio and reduction rate was considered. The calculation model of biting peak torque is built with biting time and steady-state torque influence. The model accuracy is verified by comparing the calculated data with actual data. The average deviation of steady-state torque and peak biting impact torque is within ± 8%, and ± 10%. The accuracy of these models can be improved by offline intelligent method and online learning function, subsequently.

本文基于机械动力学和轧制工艺技术，在考虑咬合冲击的情况下，进行了轧制过程中扭矩模型的研究工作。考虑到零件之间的间隙，为实际的厚板轧机的主传动系统建立了五自由度机械动力学模型。计算了不同轧制条件下的咬合峰值扭矩。分析了咬合时间和稳态扭矩的影响：咬合峰值扭矩随着咬合时间的增加而减小，并随着稳态扭矩的增加而增加。根据轧制工艺参数建立了咬合时间计算模型。通过重建水平臂比模型改进了稳态扭矩模型。考虑变形区域算术平均纵横比和缩小率的影响。建立了咬合时间和稳态扭矩影响下的咬合峰值扭矩计算模型。通过将计算的数据与实际数据进行比较，可以验证模型的准确性。稳态扭矩和峰值咬合冲击扭矩的平均偏差在±8％和±10％之内。随后，可以通过离线智能方法和在线学习功能来提高这些模型的准确性。