A Tracking State Estimator in Networks With Multirate Synchrophasor Data

This article presents a network tracking estimator using Phasor Measurement Units (PMUs) and legacy measurements. It proposes a new solution for handling PMUs with different data reporting rates. This is achieved by including the center (average or center-of-inertia) frequency as a state variable to align the PMU measurements at the estimation time. The weighted least squares estimation is formulated in complex variables via Wirtinger calculus so that the implementation can exploit Single Instruction Multiple Data (SIMD) extensions. It, therefore, allows tracking mode operation commensurate with real-time computing requirements. In addition to the conventional measurement functions, the proposed model includes new procedures for estimating the center frequency, aligning the measurements in time, and rotating phasors to initialize the tracking solution. The effect of the proposed formulation on the tracking estimator's performance metrics is evaluated on networks with up to 1888 nodes under transient conditions. The results show that properly accounting for time skew in synchrophasor data can improve the percentage error by up to four times while remaining within a 40-ms time window in successive estimation tracking.