摘要
应用MATLAB神经网络工具箱训练BP网络的基本原理和方法,建立了2 35 0中板轧机轧制压力和轧制力矩的BP网络。预报计算表明:该BP网络高效简便、计算精度好,所训练好的BP网络预报2 35 0中板轧机的轧制压力和轧制力矩的相对误差小于5 % ,满足生产控制的要求。
The BP network for calculating plate rolling force and torque in 2350 mill was established by applying the basic principle and the method of BP network trained with MATLAB network tool box.The prediction showed that the BP network was highly efficient,convenient,and with high calculating accuracy.The fractional error calculated for rolling force and torque in a 2350 plate mill was less than 5%,therefore,the method could satisfy practical production requirements.
出处
《上海金属》
CAS
2005年第4期43-45,49,共4页
Shanghai Metals