Decentralized adaptive output-feedback inverse control for a class of large-scale time delay nonlinear hysteretic systems via neural networks approximator

In this paper, an novel neural approximator based decentralized output feedback adaptive dynamic surface inverse control (DSIC) scheme is proposed for a class of larger-scale time delay systems preceded by unknown asymmetric hysteresis. The main features are as follows: 1) to our best knowledge, it is for the first time to use the neural networks and decentralized DSIC scheme to deal with both time delays and asymmetric hysteresis when only the output of each subsystem is available; 2) by combining the Finite Covering Lemma with FLSs, the assumptions of the conservative upper bound functions on the time-delay functions are removed and the Krasovskii functionals are abandoned when deal with time delays. Also, the time delay functions extends to a more general one with states variables and time-delay variables being coupled; 3) by using the initializing technique, the ℒ∞ norm of the tracking error is obtained; 4) compared with output-feedback backstepping control schemes, the controller design procedures and analysis of stability are greatly simplified by using the proposed decentralized DSIC method in which the low-pass filter at each design step is introduced. Simulation results show the validity of the proposed scheme.