论文部分内容阅读
提出一种新的递归 T- S模型 (Takagi- Sugeno模型 )的模糊神经网络结构 (TSFRNN ) ,利用动态 BP(DBP)算法来学习训练神经网络的参数 ,通过与通常的多层前馈神经网络结构的 T- S模糊神经网络(TSFNN)的对比仿真实验 ,说明在非线性系统建模方面 TSFRNN比 TSFNN更加优越 .
A novel fuzzy neural network structure (TSFRNN) of Takagi-Sugeno model is proposed. The dynamic BP (DBP) algorithm is used to study the training parameters of neural network. Compared with the usual multi-layer feedforward neural network Structure TS-NN fuzzy neural network (TSFNN) simulation shows that TSFNN is superior to TSFNN in nonlinear system modeling.