摘要
在建立模糊辨识器的一般数学模型后,利用最近邻聚类算法对样本数据进行分组,然后再将每一组数据视为一个数据对对系统进行在线训练辨识,从而使模糊辨识器能较快的收敛于真实系统。
After the normal mathematical model of fuzzy logic system is built, we can determine clusters of the sample data, then train and identify the system using the nearest neighborhood clustering algorithm. So the fuzzy identifier can converge the real system quickly.
出处
《计算机仿真》
CSCD
1997年第4期53-55,共3页
Computer Simulation
关键词
聚类学习算法
模糊系统辨识
系统辨识
Fuzzy identifier Nearest neighborhood clustering algorithm Fuzzy logic system.