Short-Term Voltage Instability Identification: A Combined Approach of Maximum Lyapunov Exponent and K-Means Clustering

This study is focused on the combination of Maximum Lyapunov Exponent (MLE) and k-means clustering for the short-term voltage instability (STVI) identification. MLE and clustering are applied to a class of differential systems that describe electromechanical and electromagnetic processes in multi-machine power grids. The authors contribute by employing this combined method for automatic labeling of voltage measurement sets, allowing the differentiation of transient regimes in the power system (PS), which can distinguish STVI among switching events in normal operational mode, faults. The obtained results can be utilized for comprehensive monitoring of transient regimes in the PS. We support our results with numerical experiments in PSCAD simulation software.