Approach of Generating Unit Model Calibration with PMU Based Problematic Parameter Identification

Abstract

Accurate model of generating unit plays a key role in power system modeling and simulation. North American Electric Reliability (NERC) requires that all generators larger than 10 MVA have to perform the model verification in every five years, to ensure the safe and stable operation of power grids. However, the key parameter identification based existing model calibration method cannot guarantee the accurate correction of the generator's problematic parameters, which have errors with their real values. The result of model calibration is still in doubt. To overcome this limitation, this paper developed a problematic-parameter-identification based model calibration method. The traj ectory sensitivity based sensitivity analysis is applied to determine the Problematic Parameter Candidates (PPCs), which have stronger influence on the model outputs to the target responses. Then, the Hilbert spectrum of PPCs and model output error curves are compared in time-frequency domain to screen out the problematic parameters. Also, the selected problematic parameters are calibrated with the measurements based identification technique. The effectiveness of the proposed method has been validated throughout a series testing cases from an actual hydropower plant integrated power system.