基于动态递归模糊神经网络的共振频率自适应反推控制  

Adaptive Backstepping Control of Resonant Frequency Based on DRFNN

作　　者：张平均[1,2] 蒋新华[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083 [2]福建工程学院电子信息与电气工程系,福建福州350108

出　　处：《信息与控制》2011年第1期21-25,共5页Information and Control

基　　金：福建省科技重点项目基金资助项目(2007H0057);福建省科技厅高校预研基金资助项目(GY-Z0880);福建省教育厅基金资助项目(JA08160)

摘　　要：针对共振破碎机频率控制系统的非线性和参数不确定性问题,提出基于动态递归模糊神经网络的自适应反推控制策略.建立了破碎机频率控制系统的数学模型,在忽略个确定性项的前提下,设计了基于自适应反推方法控制律.其次将电液比例系统中影响频率控制性能的不确定性因素定义为待估计项,采用动态递归模糊冲经网络对其进行估计,给出了基于动态递归模糊神经网络的参数自适应律,并通过Lyapunov方法证明了输出跟踪的收敛性.仿真实验和车载测试结果表明,对于参数的不确定性和负载扰动,该方法具有较好的频率控制性能.For the non-linear and parameter uncertainties of the resonant frequency controlling system of resonant machine, the adaptive back-stepping control method combined with the dynamic recurrent fuzzy neural network（DRFNN） is studied. A mathematic model of the resonant frequency controlling system is presented firstly,and the control law is designed based on adaptive back-stepping method with regardless of the parameter uncertainties.Next,the parameter uncertainties of the electro-hydraulic proportional system which affect the frequency controlling performance are defined as items to be estimated using DRFNN,the parameter adjustment law is given based on DRFNN method,and the convergence of output tracking is proved through Lyapunov function.Finally,the results from simulated experiment and test on vehicle show that this method has a better resonant frequency controlling performance for the parameter uncertainties and the load disturbance.

关 键 词：自适应反推控制 动态递归模糊神经网络 共振破碎机 破碎频率控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]