Data quality issues for synchrophasor applications Part I: a review

Abstract

Synchrophasor systems, providing low-latency, high-precision, and time-synchronized measurements to enhance power grid performances, are deployed globally. However, the synchrophasor system as a physical network, involves communication constraints and data quality issues, which will impact or even disable certain synchrophasor applications. This work investigates the data quality issue for synchrophasor applications. In Part I, the standards of synchrophasor systems and the classifications and data quality requirements of synchrophasor applications are reviewed. Also, the actual events of synchronization signal accuracy, synchrophasor data loss, and latency are counted and analyzed. The review and statistics are expected to provide an overall picture of data accuracy, loss, and latency issues for synchrophasor applications.

Data quality issues for synchrophasor applications Part II: problem formulation and potential solutions

“…In particular, the PEM algorithm employs a finite number of stored measurements for the next prediction where the store size can be chosen as small as the algorithm has a solution. Thus, different from the widely used artificial neural networks based prediction approaches which require large historical data for data modeling and training [43], the presented prediction method results in acceptable hardware cost and it is applicable for the data estimation of on-line applications [4,9,39].…”

“…The data loss issue may lower and even disable the performances of certain synchrophasor applications [4]. The incomplete or missing data can make the power grid unobservable and vulnerable, and even aggravate the cascading effects in large-scale blackouts [28], [38].…”

“…Hence, the continuous data loss will lower the accuracy of synchrophasor data and even lead to the malfunction of certain synchrophasor applications [4]. For the interpolation, it is observed that in the pre-fault and post-fault states, the lost data can be efficiently estimated (TVE \1%), whereas in the faulty state, the estimation error is acceptable only in the scenario of one or two continuous package losses.…”

“…1, with the aim of enabling wide-area monitoring, protection and control and enhancing the overall performance of power grids. To realize these great potentials, efficient measurement and communication systems are demanded [1][2][3][4][5][6].…”

PMU Data Feature Considerations for Realistic, Synthetic Data Generation

“…The unique feature sets of reported, industry PMU measurements, as observed by the consistent variations in highresolution, time-series measurements, can be attributed to several factors. Complex operation of the grid, for example, due to constantly-changing loads, control devices, operator actions, and several disturbances occurring on the system [5,6]; and the influence of malfunctioning PMU ancillary components, such as drifting internal clocks, poor global positioning system (GPS) signal reception, low-accuracy instrument transformers, improperly-connected wires, and other measurement errors lead to the familiar fluctuations observed in PMU data [7][8][9][10][11][12][13].…”