摘要
针对一类控制增益未知的多输入多输出(MIMO)非线性系统,提出了一种基于神经网络的鲁棒自适应动态面控制方法.利用动态面控制解决反推法的计算膨胀问题;同时在参数自适应律中引入S(Sigmoid)函数,动态调节神经网络的收敛速度,解决了自适应初始阶段的抖振现象.利用李亚普诺夫稳定性定理,证明了闭环系统所有信号最终有界,系统的跟踪误差最终收敛到有界紧集内.仿真结果表明了该方法的有效性.
A robust and adaptive dynamic surface control approach based on neural networks is presented for a general class of MIMO （multi-input multi-output） nonlinear systems with unknown control gain. Dynamic surface control （DSC） is used to eliminate the shortcoming of calculation explosion in traditional backstepping method. At the same time, the S- function is introduced into the adaptive mechanism so that the adaptive laws can regulate the convergence speed of neural networks, which resolve the chattering phenomenon in the initial period of adaptive control. It is shown with Lyapunov stability theory that all signals in the closed loop system are ultimately bounded and the output tracking error converges to an arbitrary small compact set. Simulation results demonstrate the effectiveness of the proposed approach.
出处
《信息与控制》
CSCD
北大核心
2008年第6期675-680,共6页
Information and Control
基金
国家自然科学基金资助项目(90405011)
关键词
自适应控制
反推法
动态面控制
RBF神经网络
adaptive control
backstepping
dynamic surface control
radial basis function neural network （RBFNN）
