一种基于RBF网络提取模糊规则的算法实现

径向基函数网络和模糊推理系统在一些柔和的情况下具有等价的功能，因此可以利用神经网络的学习算法来调节模糊系统的参数，学习后的模糊系统具有自学习和自组织性，但是削弱了模糊系统的可解释性。将模糊逻辑推理与神经网络控制技术相结合，分析了一种改进的径向基函数(RBF)神经网络结构，这种模糊神经网络结构能够有效地表达模糊系统可解释性这一突出特点，也使模糊系统具有了较好的自学习和自组织能力、通过VC 实现了基于这种RBF网络结构提取模糊规则的算法，并进行了仿真实验，仿真结果表明该算法是比较有效的。

Achievement of an Algorithm for Extracting Fuzzy Rules Based on RBF Neural Network

Radial basis function networks and fuzzy rule systems are functionally equivalent under some mild conditions. Therefore the learning algorithms developed in the field of neural networks can be used to adjust the parameters of fuzzy systems.The learned fuzzy systems are self-study and self-organize,but weaken their interpretability.Combining the fuzzy decision theory with the neural networks,and analyze an improved RBF neural network structure.This fuzzy neural network can express the interpretability of fuzzy systems,which is considered to be one of the most important features of fuzzy systems,and make fuzzy systems self-study and self-organize.Based on VC++ programming,we have achieved the algorithm of extracting fuzzy rules,and experimented.The results of simulation example show that the algorithm is an effective method.