周波数領域最小二乗近似による系統の低次線形モデル同定法

抄録

In order to perform instantaneous-value analysis of a power system, savings of computation time and memory can be achieved by reducing the complexity of external subsystems, of which the internal variables are not of interest. In this case, it is necessary to identify a reduced-order model of the external subsystem seen from its interconnection point with the main system. This paper proposes an identification method of the reduced-order model, in which a low-order linear model in the rational form of Laplace operator s is identified from frequency response data of the external subsystem using a least-squares method. The proposed method shows a good identification performance by linearization of the least-squares process combined with an iteration procedure that adaptively changes the weighting on the residue evaluation, and also by fix of gain at zero and infinite frequencies to assure the model behavior at time zero and infinity. In this paper, a five-machine radial system model is used to show the identification performance of the proposed method, and an application to control-system design is also presented.