Power system linear model selective identification by exploiting the Loewner interpolation method

摘要

This paper demonstrates the capabilities of the Loewner’s frequency interpolation method (LFIM) to extract the power system linear model in a selective-frequency range by employing the frequency response associated with inputs and outputs of the system. This is achieved taking advantage of the Loewner’s matrices that interpolate in frequency-domain the ratio between outputs and inputs. Thus, the state-space representation of the linear model is fitted providing parametric matrices, as the state matrix at one operating condition. Then, the small-signal analysis (SSA) can be addressed by computing the eigenvalues and mode-shape reproduction. Numerical results are attained after applying the proposed method to the well-known New England test system with 10 machines, 39 buses. For comparison purposes, the power system linear model is also obtained utilizing the Eigensystem realization algorithm (ERA), validating the proposition.