摘要
为了满足冷轧带钢厂对于带钢表面粗糙度预测与控制的需要,把遗传算法和神经网络结合起来形成GA-BP算法,利用遗传算法对神经网络的结构参数和性能参数进行优化,建立了一种新的冷轧带钢表面粗糙度预测模型,并应用于宝钢2030冷连轧机带钢表面粗糙度预测,预测模型具有较高的学习精度和较好的泛化能力。
In order to establish accurate prediction model of surface roughness of cold rolled strip, this paper made a combination of genetic algorithms and neural network, and it is called GA - BP algorithms , optimization of structure parameters and performance parameter of neural network was used by Genetic Algorithms . It is applied to prediction model of surface roughness of cold rolled strip in 2030 Tandem Cold Mill in Baosteel . The model has high learning precision and good generalization capability.
出处
《冶金设备》
2008年第5期42-45,共4页
Metallurgical Equipment
关键词
冷轧带钢
粗糙度
预测模型
Cold rolled strip Roughness Prediction model