| 多变量模糊推理系统的神经网络实现 |
| |
| 引用本文： | 王殿辉,柴天佑,张化光.多变量模糊推理系统的神经网络实现[J].鞍山科技大学学报,1994(2). |
| |
| 作者姓名： | 王殿辉 柴天佑 张化光 |
| |
| 作者单位： | 鞍山钢铁学院，东北大学 |
| |
| 摘 要： | 以双缸连通液面Ｆｕｚｚｙ控制系统为模型，研究了多变量Ｆｕｚｚｙ推理系统的神经元网络实现问题。这里主要考虑如下二个问题：（１）ＢＰ网络在语言环境下实现多变量Ｆｕｚｚｙ推理系统的有效性；（２）神经网络结构（隐层节点个数及输出层激发函数）对推理结果的影响。仿真结果表明，在语言环境下ＢＰ网络具有很强的近似推理能力，基于ＩＦ－ＴＨＥＮ的多变量Ｆｕｚｚｙ推理系统可由一个ＢＰ网络训练学习而加以实现，从而为有效的实现多变量复杂系统的Ｆｕｚｚｙ控制奠定了基础。
|
| 关 键 词： | 神经网络，Fuzzy控制系统，BP网络，语言环境，模糊推理 |
An Approach to Multivariable Fuzzy Inference Systems Using Neural Networks |
| |
| Wang Dianhuei,Chai Tianyon,Zhang Huaguang.An Approach to Multivariable Fuzzy Inference Systems Using Neural Networks[J].Journal of Anshan University of Science and Technology,1994(2). |
| |
| Authors: | Wang Dianhuei Chai Tianyon Zhang Huaguang |
| |
| Institution: | Wang Dianhuei;Chai Tianyon;Zhang Huaguang(Anshan Institute of I.& S.Technology)(Northeastern Universith) |
| |
| Abstract: | An approach to multivariable fuzzy inference systems using neural networks is discussed in this paper with a liquid level fuzzy contul system of two connected vessels.The main problems considered here are as follons:1) The validality of the realization of multivariable fuzzy inference systems using BP networks:2)The effect of the structure of the neural network(the number of hidden layer noder and the stimulating function of output layer on the inference results.Simulation results show that the BP network in language environment has powerful approximate inference ability and multivariable inference systems bascd on ZF-THEN statement can be realized by BP network training process,thus,an effecfive multivariable fuzzy control system is formed. |
| |
| Keywords: | neural network fuzzy control system BP network language environment fuzzy inference |
| 本文献已被 CNKI 等数据库收录！ |
