Adaptive control of interconnected systems using neural network design

In this paper, a neural network adaptive variable structure controller is proposed for decentralized control of large-scale interconnected systems. The control algorithm employs neural networks and variable structure techniques to adaptively compensate for the plant nonlinearities and for the interconnections. It is shown that the tracking error vector of all the subsystem will asymptotically converge to a neighborhood of zero and all adjustable parameters as well as the internal signals will remain bounded.