摘要
论文提出一种基于模糊逻辑和神经网络的自学习网络模型和一种结合自组织学习和BP学习的BPSOM混合学习算法。该模型通过BPSOM算法训练样本,能自动生成模糊逻辑规则,调节输入、输出变量的隶属函数;而且该算法比通常的BP学习算法收敛性好,速度快。仿真结果表明,利用该学习网络模型构造的同步发电机励磁控制器,能很好地稳定机端电压。
A learning networ k model based on fuzzy logic and neural network and a hybrid learning algorithm-E PSOM combining self-organizing learning with BP learning are proposed in this p aper.This network model can automatically produce fuzzy logic rules and adjust membership function through the EPSOM algorithm.This algorithm converges much f aster than original BP learning.Simulation results show that the synchronous ge nerator excitation controller based on the learning network can significantly st abilize terminal voltage.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第14期59-62,共4页
Computer Engineering and Applications
基金
国家教委博士学科点专项科研基金(编号:9815101)
教育部<骨干教师资助计划>项目
关键词
神经网络
自学习
模糊逻辑规则
neural network,self-learning,fuzzy logic rule
