Fast oscillation monitoring algorithms for large-scale power systems using synchrophasor data

With the installation of synchrophasors widely across the power grid, measurement-based oscillation monitoring algorithms are becoming increasingly useful in identifying the real-time oscillatory modal properties in power systems.;When the number of Phasor Measurement Unit (PMU) channels grows, the computational time of many PMU-based algorithms is dominated by the computational burden in processing large-scale dense matrices. In order to overcome the limitation, this dissertation first focuses on speeding up the Singular Value Decomposition (SVD) evaluation of large dense matrices. Many SVD problems in power system computations require only a few largest singular values of a large-scale matrix for the analysis. This dissertation introduces two fast SVD approaches recently developed in other domains to power systems for PMU-based oscillation monitoring. Both approaches are illustrated on SVD evaluation within an ambient oscillation monitoring algorithm, namely Stochastic Subspace Identification (SSI). Experimental results from archived data of the Western Interconnection demonstrate that both approaches can provide significant speedups while retaining modal estimation accuracy. With these algorithms, SVD is no longer the computational bottleneck in SSI method.;Next, this dissertation presents new formulations and computational strategies for further speeding up of oscillation monitoring algorithms. Block structures are exploited so that the large-scale dense matrix computations can be processed in parallel. This helps in memory savings as well as in overall computational time. The proposed parallel approaches are tested on three sets of archived data of the Western Interconnection with hundreds of PMU signals. Experimental results demonstrate that the real-time oscillation monitoring of the large-scale system using hundreds of PMU measurements becomes feasible.;Extra high-order is used in most measurement-based oscillation monitoring algorithms, which may bring in misleading non-physical modes. Therefore, physical mode detection (PMD) becomes an indispensable step in modal analysis. Most existing detection methods are designed for a specific modal analysis method and sometime needs empirical knowledge of the system. In the last part of this dissertation, a universal physical mode detection method is proposed utilizing frequency domain information. The new method is applied to archived data of the Western Interconnection and delivers good performance on both Prony and Matrix Pencil methods.