基于人工神经网络的热连轧参数预测模型  

Width Control Model on Finishing Train Based on BP Neural Networks

作　　者：符强[1] 张进之 

机构地区：[1]南昌大学自动化工程学院,江西南昌330029 [2]新冶高科技集团有限公司计算机与自动化事业部,北京100081

出　　处：《钢铁研究学报》2003年第4期70-72,共3页Journal of Iron and Steel Research

摘　　要：将BP神经网络的思想用于预测热连轧参数,研究并建立了基于BP神经网络的预测控制数学模型。以热连轧精轧自然宽展值为例,现场实测数据仿真验证表明,该模型明显优于传统的数学预测模型,具有很高的预测精度。In order to improve the prediction accuracy, the prediction model based on BP neural networks, which use the idea of neural net to practical engineering control model, has been designed. For width control, some good simulation results have been obtained. The BP prediction model is better than those traditional mathematic models.

关 键 词：BP神经网络 预测 精轧 宽展 

分 类 号：TG334.9[金属学及工艺—金属压力加工]