模糊对向神经网络 (FCP)在功能上同模糊逻辑系统的TS模型是等价的 ,它具有神经网络和模糊逻辑系统各自的优点 ,因而适宜作辨识模型 .利用FCP网络为辨识模型 ,设计了一种新的非线性动态系统辨识器 .辨识模型的参数调整采用无监督聚类法和误差回传算法相结合的混合学习方法进行的 ,同时文中还给出选取初始参数的方法 .仿真实验表明提出的辨识器是十分有效的 . Fuzzy Counter-oriented Neural Network (FCP) is functionally equivalent to the TS model of fuzzy logic system, and has the advantages of both neural networks and fuzzy logic systems, so it is suitable for the identification model.With FCP network as the identification model, A new nonlinear dynamic system recognizer is proposed.The parameter adjustment of the identification model is carried out by a hybrid learning method combining unsupervised clustering method and error feedback algorithm.The method of selecting initial parameters is also given in the paper.The simulation results show that The proposed identifier is very effective.