摘要
采用RBF神经网络方法建立热连轧精轧的厚度模型,通过比较有、无理论模型输入的神经网络厚度模型确定出理论数据在神经网络应用中的重要性。通过比较BP神经网络和RBF神经网络分别建立的厚度模型凸现出RBF神经网络厚度模型的优越性,并在应用过程中解决了过拟合问题。
A rolling thickness model is established using RBF neural networks. Compared with the thickness models that have or have not traditional models as input, the importance of traditional models in the application of neural networks is obvious. Rolhng thickness models are estabhshed to improve the prediction precision of roiling thickness using BP and RBF neural networks. The result of two models indicates that RBF neural networks are more accurate, and the overfitness problems in actual application have been solved also.
出处
《济南大学学报（自然科学版）》
CAS
2006年第4期312-314,共3页
Journal of University of Jinan(Science and Technology)
基金
国家自然科学基金(6057305)
关键词
人工神经网络
RBF算法
热连轧
厚度预报
artificial neural networks
RBF algorithm
hot strip mills
prediction of rolling thickness
