Decentralized adaptive stabilization of largescale interconnected linear systems

Abstract

In this dissertation, a problem of stabilizing large scale interconnected linear systems when the system parameters are uncertain is considered. A method to design a local adaptive control so that the resultant closed-loop system is assured to be globally stable is developed for both continuous-time and discrete-time systems. The proposed local adaptive controls are a combination of a new adaptive feedback control for compensation of some effects by unknown system parameters and the conventional exact model-based linear feedback control for overriding the unfavorable effects by interconnections. Based on the present decentralized adaptive method, it is shown that a class of large scale systems formed by interconnecting a number of multi-input linear systems can be stabilized via local state feedback. Both the continuous-time case and the discrete-time case are treated. It is further shown that an interconnected system composed of a number of single-input single-output linear systems can be stabilized using only input and output measurements. Both the continuous-time case and the discrete-time case are treated.