摘要
提出一种自组织模糊神经元网络控制学习方法，该方法由自组织模糊神经元网络（ＳＯＮＦ）和基于径向函数网络（ＲＢＦ）组成，具有自适应和自学习的特点．其中ＳＯＮＦ网络具有初始的网络结构与启发式模糊控制规则，能够进行结构学习与参数学习；ＲＢＦ网络用于控制对象模型的辨识，并为ＳＯＮＦ网络提供示教信号．仿真结果表明，所提出的方法控制学习效果较好．
An adaptive neuro fuzzy intelligent scheme is introduced. It consists of the self organizing neuro fuzzy (SONF) network and the Radial Basis Function (RBF) network. Before learning the initial SONF structure and heuristic fuzzy control rules are provided,an unsupervised learning algorithm is applied to structure learning if necessary. The BP algorithm is used to adujst the nodes and links. So the combination of these two algorithms can partition the input/output space in a flexible way based on the distribution of the training data. The RBF network is used for system identification,providing the SONF network with the teaching signal. The proposed scheme has two important characteristics of adaptation and learning. Computer simulations present the performance and applicability of the proposed scheme on injection system of blast furnace.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1998年第4期362-365,共4页
Journal of Northeastern University(Natural Science)
基金
国家计委"九五"科技攻关项目
关键词
模糊控制
自组织学习
模糊神经元网络
fuzzy control,neural network,self organizing learning,neuro fuzzy network.
