本文对一类用输入输出模型表示的非线性系统,基于模糊神经网络,提出了一种滑模自适应控制方法.通过对系统输入输出模型的分解,将控制器的设计分为两步,第一步是设计模糊神经网络滑模控制器,第二步是进行线性系统的设计.模糊神经网络滑模控制器是由模糊神经网络实现滑模控制,它平滑了切换信号,消除了滑模控制中固有的颤动现象同时在控制器的设计中不需要知道系统的不确定性和扰动的界限,使系统有强鲁棒性.运用Lyapunov稳定理论,证明了整个系统是稳定的且系统的跟踪误差收敛到零的邻域.仿真结果表明了控制方案的有效性.

Based on Fuzzy Neural Networks (FNN),a sliding mode adaptive control methodology is presented for a class of nonlinear systems represented by input-output models.There are two parts in the controller design:one is the design of the Fuzzy Neural Networks Sliding Mode Controller(FNNSMC) and the other is the design of the linear feedback system.The FNNSMC,where the FNN is used to fulfil the sliding mode control,can eliminate the chattering by smoothing the control signal while the bounds of the uncertainties and the disturbances of the systems are not known in the controller design.By the Lyapunov's stability theory,we have proved that the system is globally stable and the tracking error can be converged to the neighborhood of zero.