An improved Stochastic Subspace Identification based estimation of low frequency modes in power system using synchrophasors

Abstract

Accurate estimation of low frequency modes in the power system is vital for improving its small signal stability. Stochastic subspace identification (SSI) is a subspace based method which provides fairly accurate estimates of these low frequency modes under normal operating conditions. However, SSI based methods require prior information about the number of frequency components present in the signal and its accuracy of estimation degrades under noisy conditions. To overcome these drawbacks, an improved SSI method using a combination of Stationary Wavelet Transform (SWT) and Exact Model Order (EMO) algorithm for estimating the modal parameters of the low frequency oscillations is proposed in this paper. SWT is used for denoising the signal thereby improving the noise resistance whereas EMO is used for estimating the model order accurately to reduce the computational complexity of the SSI method. The proposed method is tested using synthetic signals as well as the PMU data from two practical power systems. The validity of the proposed method has been verified at different levels of noise. The comparison of the proposed method with Fourier, Teager Kaiser, Matrix Pencil, Eigen Realization Algorithm and SSI based methods reveal its accuracy and robustness even under high noise contamination. © 2019 Elsevier Ltd