具有逆辨识结构的模糊神经网络控制及其应用  

Control and Application of the Fuzzy Neural Network with Inverse Identification Structure

作　　者：吴学礼[1] 孟华[2] 孟凡华[2] 王永骥[1] 李平[2] 贾辉然[2] 

机构地区：[1]华中科技大学控制科学与工程系,湖北武汉430074 [2]河北科技大学电气信息学院,河北石家庄050054

出　　处：《地理与地理信息科学》2004年第5期110-112,共3页Geography and Geo-Information Science

基　　金：河北省科技攻关计划项目 ( 0 12 13 40 7D)

摘　　要：该文提出一种新型模糊神经网络结构及算法。在这种控制方案中 ,采用三层模糊神经网络控制器和神经网络逆辨识控制器相结合的结构。计算机仿真研究和实际应用表明 ,采用新型模糊神经网络控制方法 。The model reference adaptive control based on fuzzy neural network is generally adopted.Because the online adjusting of fuzzy neural network controller need neural network as error back propagation passageway,so the number of layers of the controller and identification is too big and it is easy to cause error back propagation uncertain,as a result,the reference of the controller can't be modified accurately.Besides,this method makes the controlled system long adjusting time and certain steady-state error.This paper proposes a new fuzzy neural network structure and algorithm.In this scheme it has the structure that combines a three-layer fuzzy neural network controller with neural network inverse identification controller.Computer simulation and actual application show that this scheme has effective control to larger time-delay nonlinear system.

关 键 词：逆辨识结构 模糊神经网络控制 联想记忆神经网络辨识 电阻加热炉温控系统 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]