Decentralized Adaptive Control Of Interconnected Systems

Large-scale system is a complexity and important system. Nowadays, more and more people have paid attention to the study of large-scale system. Owning to the complexity of control synthesis and the physical restrictions on information exchange it is often required to design a decentralized controller depending only on local measurements for each subsystem, even if to achieve an objective for the whole large-scale system. Compared with conventional large-scale system, strictly decentralized adaptive control has been studied well. However, It is worth pointing out that most of previous literatures have not dealt well very, or not dealt with the problem of time-delay. It is well known that time delay commonly exists in the practical systems and it may destabilize the control systems. In this thesis, we focus upon the problem of strictly decentralized adaptive control for large-scale interconnected systems. The main results are as follows:1) Strictly decentralized adaptive control for a class of interconnected systems.we address the problem of strictly decentralized adaptive control for a class of interconnected systems. Three types of interconnected systems are introduced. For subsystems whose interconnections are realized by state variables, we consider linear and nonlinear interconnected systems simultaneously; For subsystems that output, rather than state variables affect the performance of other subsystems, we assume the transfer functions of all the subsystems have unity relative degrees. Some decentralized adaptive control scheme, capable of guaranteeing the asymptotical tracking with zero tracking error, is proposed. Finally, a numerical example is provided to demonstrate the effectiveness of the main results.2) Decentralized model following adaptive control for a class of large-scale interconnected systems with multiple time-varying delays.The control problem is addressed for a class of nonlinearly interconnected systems with time-varying delay and disturbances. Two types of time-varying delays are considered simultaneously. For the time-varying delays that we assume their derivatives have to be less than one, we can prove that the tracking errors are uniformly ultimately bounded; For the time-varying delays whose derivatives don’t be required less than one,we can prove that the tracking error uniformly exponentially converge towards a ball which can be as small as desired. Finally, a numerical example is provided to show the feasibility and validity of the developed results.