Synchrophasor based dynamic model validation leveraging multiple events

摘要

This paper proposes a method to validate and calibrate power system dynamic models using measurements from multiple events. By simultaneously calibrating the models across multiple events, the resulting parameter values can rightly explain the model response across all these events. The proposed solution has three components: (i) parameter identifiability analysis, (ii) event characterization, (iii) parameter estimation. The identifiablity analysis accounts for multiple events simultaneously, thereby providing a global quantification of parameter identifiability. The event characterization approach clusters similar events and determines the set of most representative events from the available measurements. Finally, an Unscented Kalman Filtering based parameter estimation algorithm is adopted to tune the model parameters across multiple events simultaneously. The efficacy of the proposed solution is demonstrated using a playback simulation platform based on the Matlab based power system toolbox.