论文部分内容阅读
德国赫施钢公司，原先以传统的数学模型对一台宽带钢轧机各项目标值与热轧过程技术特性曲线进行的预调，对于工艺过程所作的描述既不全面又不精确，具体调整时还须重新计算各项模型参数，才能适应现行的工艺操作。鉴于上述，西门子公司为该厂开发了采用神经网络的预调系统。该
Germany Hearst steel company, the original mathematical model of a traditional wide strip mill of the target value and the hot-rolled process technical characteristics of the preset curve, the description of the process is neither comprehensive nor accurate, the specific adjustment The model parameters must be recalculated in order to adapt to the current process operation. In view of the above, Siemens has developed a pre-set system using neural networks for the plant. This