Robustness to missing synchrophasor data for power and frequency event detection in electric grids

Phasor Measurement Unit (PMU) data provides time synchronized measurements of important electrical and power signals in AC power grids. The sheer volume of three phase electric grid PMU data, typically measured at 60 samples second, necessitates the use of automatic event (anomaly) detection for grid monitoring and control. However, this becomes an unwieldy task in cases of PMU data dropouts due to the unobservable state of the grid. Hence, robustness to missing phasor measurements would be critical in order to monitor and control the electric grid in an orderly manner. This paper illustrates an approach for interpolation of PMU data in case of missing data points by optimal filtering of phasor data along with event detection. It is shown that optimal filters can be estimated on the basis of non linear recursive search optimization on real-time PMU data and these filters can be used to generate forecasts in case of missing PMU measurements. The approach is illustrated on phasor data obtained from a microPMU system developed by Power Standards Lab for data ride-through in case of dropouts.