摘要
在一定的前提条件下,提出一种简化的模糊RBF网络模型.该网络由输入层、模糊化层、模糊联结层、合成联结层和输出层组成.同时,还给出网络的构造方法和自学习方法.仿真结果表明此网络对非线性、多变量函数具有良好的逼近能力.
If fuzzy neural network is constructed by prototype RBF network directly, the size of network will be too large to apply when the system is multivariable. A new simplified model of fuzzy RBF networks is presented based on the functional equivalence between the fuzzy inference system and the radial basis function (RBF) network. The self construction method and self learning algorithms of this simplified model are discussed. Finally, a computer simulation is given. The simulation results show that the fuzzy network simplified model has a quite satisfied approximability to a nonlinear and multivariable function.
出处
《华中理工大学学报》
CSCD
北大核心
1998年第9期30-33,共4页
Journal of Huazhong University of Science and Technology
基金
广东省自然科学基金
关键词
模糊逻辑
神经网络
人工智能
智能控制
RBF网络
fuzzy logic
neural network
artificial intelligence
intelligent control
function approximation