Adaptive Neural Control Of Several Classes Of Nonlinear Systems With Unmodeled Dynamics

Based on the related researching findings on uncertain nonlinear systems,this thesis will apply neural control theory,adaptive backstepping design and robust control theory to study the control problems of nonlinear systems with unmodeled dynamics and propose effective adaptive neural control approaches.The compedious frame and statements of this thesis include the following two parts:1.For a class of non-affine nonlinear time-delay systems with dynamic uncertainties,we develop an adaptive neural intelligent control scheme.In the design of the controller,a variable partition technique is applied to surmount the difficulty in the nonlinear functions of whole state variables.By using the backstepping recursive design approach and the universal approximation capability of neural networks.An adaptive neural controller is systemically designed to ensure that all signals of the close-loop system are finally bounded.The simulation results verify the effectiveness of the proposed control scheme.2.The adaptive neural network control problem for strictly feedback nonlinear system s w-ith dynamic uncertainties and input saturation is considered.In the control design,Ridial basis function(RBF)NN are employed to model the unknown nonlinearities,and an adaptive neural control schene guarantees all the signals in the closed-loop system are semiglobally uniformly ultimately bounded in mean square.By estimating the maximum value of the neural network's weight vector norm,the control system only needs an adaptive parameter,which makes the computational burden alleviated.A simulation example is provided to show the effectiveness of the proposed control scheme.