摘要
以4200轧机大规模测试得到的实验数据为基础,利用Matlab人工神经网络工具箱,建立了该轧机轧制力的预测模型。通过对该网络隐层神经元个数的调整,提高了收敛速度,使轧制力的预测精度大为提高。
Based on the data obtained from large scale experiments, the prediction model of this mill was built. The convergence speed was inereased and the prediction accuracy of the draught pressure was improved by adjusting the number of hidden--layer neurons in the network.
出处
《钢铁研究学报》
CAS
CSCD
北大核心
2007年第6期95-98,共4页
Journal of Iron and Steel Research
基金
国家自然科学基金资助项目(50175031)
关键词
轧制力
BP网络
预测
draught pressure
BP neural network
prediction
