Online Transient Stability Assessment Using Regression Models

Abstract

AbstractOnline transient stability assessment (TSA) is very much important and much needed for quick detection of transient instability triggered by different contingencies. An online transient stability assessment approach based on the prediction of transient stability margin using regression models is proposed in this paper. A single-machine equivalent (SIME)-based accelerating power is used for defining a normalized transient stability margin. For training regression models, voltage magnitudes of severely disturbed generator measured at four different instances and pre- and during fault One-Machine Infinite Bus (OMIB) rotor angles are used as inputs. Accelerating power-based normalized transient stability margin is the output of the regression models. The models are trained offline for different load conditions. These trained models are used online to predict the stability margin which in turn gives information about the post-fault transient stability status. The proposed approach is applied to the New England 39 bus test system and the results are validated with time-domain simulation results.KeywordsSingle machine equivalent (SIME)Time-domain simulation (TDS)Transient stability assessment (TSA)Transient stability margin (TSM)