摘要
提出了一种应用于热工过程在线建模的改进的RAN网络。该网络利用“新息准则”对隐层节点进行精调 ,减小了新样本对已学习好的网络参数的影响 ,并避免了过拟合。还分析了网络初始参数对网络结构的影响 ,并给出了这些参数的合理取值范围。仿真研究表明 ,改进的RAN网络用于热工过程在线建模时收敛速度快 ,泛化能力好 ,逼近精度高。
An improved RAN(Resource-Allocating Network)for thermal processes on-line modeling is presented in this paper. The improved RAN can suppress the interference of new data to the connection weights trained formerly. The influence of preliminary parameters on the network is also illustrated and property values of them are given in this paper.Simulation results show that the improved RAN has rapid learning rate,good generalization ability and high approximation accuracy.
出处
《锅炉技术》
北大核心
2004年第5期12-15,63,共5页
Boiler Technology
