中厚板轧制过程智能化温度预报模型

 采用有限元(FEM)程序模拟计算了中厚板轧制过程中的温度变化，得到与实测温度符合甚好的模拟结果。以模拟计算结果为基础，建立了BP神经网络和回归温度预报模型。采用两种模型对中厚板热轧过程中轧件表面温度变化情况进行了预报。结果表明，神经元网络模型的预报值较回归模型更接近FEM模拟计算值和实测值，可将神经元网络模型应用于中厚板轧制过程中轧件表面温度变化的在线预报。

Intelligent Temperature Prediction Model for Plate Rolling

The temperature variation during plate rolling was simulated by using FEM. The simulation results were in good agreement with the measured values. BP neural network and Regression models were built based on the simulation results, and predictions about temperature variation were carried out by using the two models. The results showed that the neural network model was more accurate than the regression model, indicating that the neural network model can be applied to on-llne temperature prediction during plate rolling.