摘要
针对系统的非线性多变量耦合特性 ,提出了一种基于隶属函数的模糊神经网 (MFFNN)智能解耦的新算法 ,通过神经网络的自学习能力 ,调节隶属函数的形状及结论值 ,使 MFFNN具有自学习、自适应的能力 ,并用主机、接口及炉群组成系统验证了该算法 ,仿真及用于炉群智能解耦实测数据表明 。
According to the coupling characteristic of nonlinear multivariable, a new algorithm of intelligent decoupling of fuzzy neural network based on membership function (MFFNN) is presented.The neural network's self learning and self adapting ability can adjust the membership function shape and the conclusion value to make the MFFNN also has the ability of self learning. The stoves' decoupling control is tested by this algorithm.The simulation result and the experiments show that this intelligent algorithm has a good performances of eliminating the coupling of the nonlinear multivariable.
出处
《西安公路交通大学学报》
CSCD
北大核心
2000年第4期123-125,129,共4页
Journal of Xi'an Highway University
关键词
多变量耦合
隶属函数
模糊神经网
智能解耦算法
multivariable coupling
membershis function
fuzzy neural network
intelligent decoupling algorithm