一种基于神经元的自适应模糊推理网络

本文根据Ｔ－Ｓ模糊模型提出了一种新的基于神经元的自适应模糊推理网络，给出了连接结构和学习算法，它能自动学习和修正隶属函数及模糊规则，将其用于Ｂｏｘ的煤气炉，太阳黑子预报以及降雨量预报等不同类型的复杂系统建模，仿真结果表明，该模糊神经网络具有收敛速度快，辨识精度高，泛化能力强和适应范围广等特点，可当作复杂系统建模的一种有效工具。

An Adaptive Fuzzy Neural Inferring Network

In this paper,a 'new adaptive fuzzy neural inferring network ased on the T-S model ispropose. Also, the network structure and the learning algorithms are given. The fuzzyneural inferring network is capable of learning and modifying the membership functions andfuzzy rules by itself,and it is used to model several complex systems such as the gas furnacedata of Box and the sunspots data. sa will as raifall data etc. The simulation results show thatthe network has the properties of fast convergence, high accuracy and better capability ofgeneralizing as well as broad field of utilizing. So it is a powerful tool for modeling complexnonlinear systems.