Decentralized direct adaptive neural network control of interconnected systems

The problem of direct adaptive neural network control for a class of large-scale systems with unknown function control gains and the high-order interconnections is studied in This work. Based on the principle of sliding mode control and the approximation capability of multilayer neural networks, a design scheme of decentralized direct adaptive sliding mode controller is proposed. The plant dynamic uncertainty and modeling errors are adaptively compensated by the adjusted weights and sliding mode gains on-line for each subsystem only using local information. According to the Lyapunov method, the closed-loop adaptive control system is proven to be globally stable, with tracking errors converging to a neighborhood of zero.