Neural-network-based decentralized output-feedback control for nonlinear large-scale delayed systems with unknown dead-zones and virtual control coefficients

Abstract

An adaptive decentralized output-feedback control problem is investigated for a class of pure-feedback large-scale nonlinear systems with mismatched uncertainties, unknown dead-zones and unknown virtual control coefficients. At the same time, the nonlinear interconnections involve time-varying delays. Because only output information can be obtained, a state observer is constructed first. By using the infinite approximation ability of the radial basis function neural networks, the difficulty caused by the unknown nonlinearities is successfully overcome. Based on the appropriate Lyapunov–Krasovskii functions, the time-delay terms are compensated. The unknown virtual control coefficients are disposed by the convex combination method. By combining the backstepping technique with decentralized control principle, the adaptive neural decentralized output-feedback controllers are constructed to guarantee all the signals of the resulting closed-loop systems are bounded. And meanwhile, the error signals can converge to a small neighborhood of the origin. The simulation examples are provided to test our results.