摘要
在分析轧机轧制过程中的力能参数的基础上 ,测量并计算了 50 0 mm轧机的轧制力。探讨了用神经网络模型确定轧制力的方法 ,预测了轧制力 ,并将三种方法得到的轧制力进行了比较。结果表明用神经网络模型预测的轧制力误差远小于基于传统数学模型计算的轧制力误差 ,它能较好他反映实际轧制过程的特征。
Based on analysis of mechanical parameters of rolling mill in rolling process, the rolling force of 500mm rolling mill was tested and calculated. The method for setting up rolling force with the neural network was discussed. The rolling force was predicted and three kinds of rolling force obtained from testing, calculating and modeling of neural network were compared. The result shows that the error of rolling force by neural network is less than the calculation error based on convenience mathematical model,which fairly reflects the real feature of the rolling process.
出处
《重型机械》
2000年第4期35-37,共3页
Heavy Machinery
关键词
轧制力
神经网络预测
力能参数
轧机
rolling mill, rolling force, neural network, prediction
