Intelligent Power System Stability Assessment and Dominant Instability Mode Identification Using Integrated Active...

Area of Science:

• Electrical Engineering
• Computational Intelligence
• Power Systems Analysis

Background:

• Power system stability analysis is crucial for operational security.
• Rotor angle and voltage stability are often intertwined, complicating hazard identification.
• Current dominant instability mode (DIM) identification relies heavily on human expertise.

Purpose of the Study:

• To develop an intelligent framework for automated DIM identification.
• To discriminate between stable, rotor angle unstable, and voltage unstable power system states.
• To reduce the reliance on manual expert labeling for deep learning model training.

Main Methods:

• Implementation of an active deep learning (ADL) framework.
• Design of a two-stage batch-mode integrated ADL query strategy (preselection and clustering).
• Focus on selecting the most informative and diverse samples for labeling to enhance query efficiency.

Main Results:

• The proposed ADL framework accurately discriminates between stable, rotor angle unstable, and voltage unstable conditions.
• The two-stage query strategy significantly reduces the number of labeled samples required.
• The approach demonstrates superior accuracy, label efficiency, scalability, and adaptability compared to conventional methods.

Conclusions:

• The intelligent DIM identification framework offers an effective, automated solution for power system stability assessment.
• Active deep learning with an efficient query strategy minimizes expert labeling efforts.
• The method provides a robust and scalable approach for ensuring power system security.