ADAPTIVE DECENTRALIZED SLIDING MODE NEURAL NETWORK CONTROL OF A CLASS OF NONLINEAR INTERCONNECTED SYSTEMS

In this paper, a completely Decentralized control method (DNNS) for a class of large-scale interconnected systems is developed based on the combination of the sliding mode control with the Neural Network (NN). The standard sliding mode control (SMC) can be used; however, for systems with unknown interconnection terms and in the presence of large uncertainties, the result controller is with higher switching gain and introduces higher amplitude of chattering, or may completely diverge. In this study, the NN is used to predict the unknown interconnection terms and the unknown part of model for each subsystem, and hence it enables a lower switching gain to be used. The stability is shown by the Lyapunov theory and the control action used did not exhibit any chattering behaviour. The effectiveness of the designed DNNS is illustrated in simulations by a comparison with standard SMC technique.