Indices for Automated Identification of Questionable Generator Models Using Synchrophasors

Traditional methods for validating generator models require taking generating units offline and performing staged tests. Techniques that are developed using synchrophasor data from Phasor Measurement Units (PMUs) allow validation of generator models without taking units offline thereby saving cost and avoiding potential damage to equipment. Typically, identifying whether a model is questionable requires a visual comparison of simulated response with actual measured response from PMUs for grid events and disturbances. This paper proposes a normalized root mean square deviation index and combines it with existing indices to automatically assess the validity of a model. The indices can quantify the mismatch between model response and actual measured response comprehensively and accurately. Additionally, an algorithm for identifying the fault period of an event is proposed in order to accurately compare the model and actual measured responses using the proposed new indices. Case studies verify the effectiveness of the proposed indices with actual PMU measured events and artificially created cases.