摘要
基于原有励磁建模方法,将模糊理论和神经网络相结合,充分利用模糊理论处理不精确问题和神经网络较强的泛化能力等优势,形成一种模糊神经网络(FNN),并推导了FNN的学习算法.将此FNN用于发电机的控制环节——励磁系统建模.仿真结果显示,FNN模型能够较精确地对实际系统进行拟合.
Precise model plays a vital role in simulation of dynamic process of power systems. By combining the fuzzy theory and neural network technology, a fuzzy neural network (FNN) was proposed, whose learning algorithms are developed by steep algorithm. The excitation system model based on FNN was also derived, which can be used for on-line and off-line analysis and control respectively. The simulation results demonstrate that the FNN model can give precise estimation of practical excitation system.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2005年第S1期5-8,共4页
Journal of Shanghai Jiaotong University
关键词
励磁系统
系统辨识
模糊神经网络
excitation system
system identification
fuzzy neural network （FNN）
