【摘要】 在S Stoeva提出的基于相同样本及网络输出的模糊反向传播算法基础上,通过对基于极大-极小模糊算子的模糊神经元模型的研究,对含有一个隐含层的单输出模糊神经网络,提出了依赖于各模糊神经元输出的调整模糊权值的网络学习算法,该算法具有直观和可操作性强的特点.并以汽轮发电机组的状态监测为例进行仿真,仿真结果表明网络学习效果较好.更多还原

【Abstract】 The paper studied the fuzzy neurons model for max-min fuzzy operator based on S.Stoeva’s fuzzy backpropagation algorithms.Then it proposed the networks learning algorithms depended on fuzzy neurons output to change fuzzy power for fuzzy neural networks of only output and implication layer.The algorithms have characteristics of intuition and operation.Finally,the paper simulates experiment such as state monitoring of turbo-generator set,and shows that the networks learning results are good.更多还原