Adaptive output feedback tracking for a class of large-scale nonlinear systems with measurement uncertainties

In this paper, a partially decentralized adaptive output feedback tracking scheme is proposed for a class of strongly coupled large-scale nonlinear systems for which it is assumed that both multiplicative and additive measurement uncertainties are taken into consideration, and the system local nonlinearities and nonlinear interactions include not only the outputs of the subsystems but also their internal states. A dynamic gain shared by all the local state observers and controllers is designed, with which the effects of the unknown measurement uncertainties can be adaptively compensated despite uncertain parameters, local nonlinearities, nonlinear interactions and external disturbances. It is proved that each local controller can be designed to be linear-like, which, together with the dynamic gain, can greatly simplify the controller design of the large-scale system. By introducing some coordinate transformations and properly choosing Lyapunov functions, it is proved that all the closed-loop signals are bounded and the relations between the tracking errors (real and ideal tracking errors) of each subsystem and the corresponding measurement uncertainties can be explicitly established.