论文部分内容阅读
该文提出了一种基于规则的Ｔ－ Ｓ模糊神经网络的结构和相应的算法。首先用自组织算法对学习数据进行聚类生成一组初始的模糊规则，然后用误差反传法细调网络参数，通过仿真验证，该模糊神经网络具有结构简单，拟合精度高等优点。
This paper presents a structure and corresponding algorithm of rule-based T- S fuzzy neural network. Firstly, self-organizing algorithm was used to cluster learning data to generate a set of initial fuzzy rules. Then the error parameters were used to fine-tune the network parameters. The simulation results show that the fuzzy neural network has the advantages of simple structure and high fitting accuracy.