Decentralized adaptive output feedback dynamic surface control of interconnected nonlinear systems with unmodeled dynamics

This paper is concerned with the centralized and decentralized adaptive output feedback control problem for a class of uncertain interconnected nonlinear systems with unmodeled dynamics and unmeasured states. The known requirement of the upper bounds of dynamic disturbances and coupling terms in the considered system is avoided. Decentralized observers are constructed based on the K-filters to estimate the unmeasured states of each subsystem. In the design of adaptive output feedback dynamic surface control, the unknown nonlinear functions are directly approximated by neural networks, while the coupling terms being lumped together with those unknown continuous functions produced in deduction are approximated as a whole. Using the defined compact set in the stability analysis, some unknown nonlinear continuous functions are dealt with effectively. It is proved that all the signals in the closed-loop system are semi-globally uniformly ultimately bounded. Simulation results are provided to illustrate the effectiveness of the proposed approach.