Short-Term Voltage Stability Prediction for Power Systems Based on a Dominant Koopman Operator-Enhanced MLE

Abstract

Short-term voltage stability (STVS) prediction is a critical technology for modern power systems with high penetration of renewable energy resources. To address the limitations of the traditional maximum Lyapunov exponent (MLE) in handling short-time window data during transient dynamics, this paper proposes a novel dominant Koopman operator-enhanced MLE (DKO-MLE) method for STVS prediction. First, the dominant Koopman operator is extracted using a proposed approach that combines the Hankel matrix with the extended dynamic mode decomposition (HEDMD) method. Subsequently, the extracted dominant Koopman operator is predicted in the time domain and utilized for MLE calculation. The proposed method leverages the data dimensionality expansion capability of the Hankel matrix and the time-domain prediction ability of the dominant Koopman operator to effectively generate sufficient data for online STVS prediction. Case studies conducted on the Nordic32 system and a standard test system provided by the China Electric Power Research Institute (CEPRI) demonstrate the accuracy and rapidity of the proposed method in predicting short-term voltage stability.

Keywords

• Power systems
• maximum Lyapunov exponent
• short-term voltage stability
• dominant Koopman operator