摘要
基于神经网络原理 ,采用 BP算法训练网络 ,建立热轧控制参数 (轧制温度 ,化学成分 ,变形量等 )对描述产品力学性能的参数的映射关系。离线仿真表明 ,将神经网络模型应用于热轧控制预报 ,具有一定的现实意义。
This article makes a brief description of BP neural network model,adopts BP algorithm training network based on neural network principle,establishes an alluding relation with respect to the parameters(rolling temperature,chemical composition,amount of deformation,etc)for controlling hot rolling and the parameters(yield strength,tensile strength,elongation)for expressing mechanical properties of product.Off-line emulation shows that application of neural network model to hot rolling control forecast will have a certain real significance.
出处
《宽厚板》
2003年第6期11-14,共4页
Wide and Heavy Plate
