Power system transient stability feature filter based on grey incidence clustering

Transient stability analysis is a highly complex nonlinear problem in large-scale power systems. Lots of features can reflect the transient stability state. Selecting the appropriate input features which have strong relations with stability or stability index is one of the key factors that acquire a better result in the transient stability evaluation. This paper presented a method that excellently combined the grey incidence optimizing order with the clustering analysis. This method can not only get different results of clustering analysis for various threshold values, but also be used in further analysis such as choosing the optimal features or optimal representations in a congeneric class. The simulation result demonstrated that the method was effective and correct. This paper provided a new idea for the transient stability evaluation feature filter in large-scale power systems.