Adaptive Dynamic Surface Control For Uncertain Nonlinear Systems

Backstepping control technology has been widely used in adaptive control for nonlinearsystems in recent years. Dynamic surface control approach is introducing first-order filteringbased on the backstepping method. This approach can avoid tedious derivation and lower thecomplexity of design procedures. To take advantage of the approximation ability of neuralnetworks can solve the controller design problems of nonlinear systems with unknown functions.This paper studies the controller design problems of uncertain nonlinear systems via combiningthe above methods. It mainly involves the following several aspects: adaptive control; dynamicsurface control; output feedback; neural network control; unmodeled dynamics; Lapunovstability. The main work of this paper is as follows:The problem of a class of nonlinear systems with unknown parameters and unmeasuredststes is considered in this paper. To solve this problem, an adaptive output feedback controlalgorithm based on the dynamic surface control is proposed. With the approximation ability ofneural networks, a controller is designed based on the dynamic surface control respectively forthe problems of known and unknown boundary parameters. Both schemes can effectively avoidthe complexity caused by a series of differentiation of the virtue control. Simulation studies areconducted to verify the effectiveness of the proposed approach.A effective approach is presented for a class of completely non-affine pure-feedbacknonlinear systems with unmodeled dynamics. The radial basis function neural networks are usedto approximate the uncertain nonlinear functions obtained in the reference procedure.Only oneparameter is needed to be updated in this scheme. This method can avoid the complexity causedby a series of differentiation of the virtue control. Simulation studies are conducted to verify theeffectiveness of the proposed approach.The controller design and stability analysis of several classes of uncertain nonlinear systemshave been solved in this paper. And the simulation examples of each scheme verify theeffectiveness of the proposed approach.