Online Power System Security Analysis and State Prediction

In this Dissertation, models of power system components for different system states from normal to power system restoration are proposed and their functionalities are examined. These models are used for different prototypes which could be used at control centres. The static equivalent models of ADNs are proposed which enhance the quality of the SSA when compared to the existing passive load modelling solution. These models address both conditions of having full or lack of ADN observabilities. A gradient-based probabilistic vertical load model is proposed which can be utilized in a probabilistic SSA tool. It is developed based on the historical vertical load measurements available at control centres. As the vertical loads might be correlated and their correlations might vary by time or RES generation levels, a scenario-based probabilistic SSA is proposed. A simple optimization tool is proposed which assist the DSOs to segregate their networks into load clusters which are desired by the TSOs. This optimizer requires models of the network components during the system restoration.