双输入型模糊前向神经网络的构建  

Construction of Dual-Input Fuzzy Feedforward Neural Network

作　　者：杨文光[1] 闫守峰[1] 文小艳[1] 

机构地区：[1]华北科技学院基础部,中国北京101601

出　　处：《湖南师范大学自然科学学报》2013年第3期33-38,共6页Journal of Natural Science of Hunan Normal University

基　　金：华北科技学院高等教育科学研究资助课题(HKJYZD201213);中国高校基本科研业务费资助项目(3142013021)

摘　　要：利用高斯型隶属函数作为隐层神经元激励函数,构建了四层模糊前向神经网络.根据从训练数据集中提取出的插值样本数量来确定隐层神经元个数.网络结构确定后,基于二元函数逼近论确定最优权值,得到双输入型近似插值神经网络,说明了最优权值的双输入型模糊前向神经网络的实现过程.计算机数值仿真实验表明所构建的网络在运行时间、逼近精度与去噪效果等方面是有效的,丰富了多输入神经网络的构建方法.By the use of the Gauss type membership function as the hidden layer neuron activation function,the four layer fuzzy neural networks were constructed.According to the extracted training data set of the interpolated samples,the number of the hidden layer neuron was determined.Firstly,the network structure was constructed.Based on the optimal weight from binary function＇s approximation,the dual-input type of approximate interpolation neural networks was received.Then,the realization process of the optimal weight dual-input network was explained.Finally the computer simulation results show that the constructed network is effective at run time,the approximation accuracy and de-noising effect,and multi input neural network construction method is richened.

关 键 词：前向神经网络 高斯型隶属函数 权值直接确定 插值 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]