Decision tree-based classification of multiple operating conditions for power system voltage stability assessment.

• A classification criteria that enlarges and generalizes security boundary methods. • It classifies the operating space based on the nearest Saddle Node Bifurcation (SNB). • A sampling method that reduces computation burden by exploiting SNB conditions. • An algorithm that creates a database of operating points to train decision trees. • A testing method for decision trees accuracy prediction using time-domain simulation. This paper presents a method that performs classification of thousands of operating conditions w.r.t. power system voltage stability by using decision trees. The proposed method uses a new and flexible classification criterion that allows to identify operating conditions that are near or within the region for which the system is voltage unstable, and more importantly, that can consider operational requirements. The method creates both training and test data sets when building and validating the decision trees. To minimize computational burden, a sampling method is proposed, this method exploits the Saddle Node Bifurcation conditions to explore the operational space used to train the decision trees. Case studies were performed using the IEEE 9-bus system for several operating conditions and different network configurations. This paper also proposes the use of time domain simulations to assess the prediction accuracy of decision trees. Decision trees were created for network configurations involving outage of the line were tested on test sets and also using time domain simulations results from PSS/E. The ability to classify the degree of voltage stability of a multitude of operation conditions could be useful to aid operators in selecting and applying preventive measures to steer away the system from unstable conditions or conditions that are close to breaching operational requirements w.r.t. voltage stability. [ABSTRACT FROM AUTHOR]