Decentralized Model Reference Adaptive Control of Multi-Machine Power System

Abstract

The researches on decentralised and/or adaptive control of power systems have received a great interest in recent years. In particular, the emphasis of most adaptive approaches towards improved stability has been placed on the utilisation of controller based on single-input/single-output models of single machine systems. Unfortunately, this approach does not guarantee an adequate performance in multi-machine system due to the strong interaction among machines. For this reason, in the paper, a decentralised model reference adaptive control (DMRAC) technique has been used to design a controller for a multi-machine system. In fact, considering as subsystem reference model for each machine the dynamic system determined by an appropriate optimal control law, the proposed DMRAC allows to adjust on line regulator gains in order to compensate the fluctuations in the interconnection. In order to show the performance of the proposed control law, some numerical experiments have been made on a four machines test system.