摘要
针对BP神经网络在学习速度方面的不足,在Jordan和Elman网络结构的基础上,提出了一种带偏差单元的IRN(internallyrecurrentnetwork)网络模型,根据BP算法推导出了该网络模型的权系数调整规则,并应用该网络模型进行了故障诊断方面的仿真分析.试验结果表明,该网络模型的收敛速度比一般BP网络有了很大提高,具有很好的实用性.
To deal with the weakness of the BP neural network in learning speed, an internally recurrent network model with bias cells is presented based on the Jordan and Elman neural networks. The weight-regulating method is developed based on BP algorithm. Simulations on fault diagnosis are performed with this neural network model. Experimental results show that the converging speed of this network model is faster than the traditional BP network and this model has a good practicability.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2005年第5期399-401,共3页
Transactions of Beijing Institute of Technology
关键词
BP神经网络
递归神经网络
故障诊断
系统仿真
BP neural network
recurrent neural network
fault diagnosis
system simulation
