Adaptive non-parametric kernel density estimation for under-frequency load shedding with electric vehicles and...

Area of Science:

• Electrical Engineering
• Power Systems Engineering
• Renewable Energy Integration

Background:

• Increasing integration of variable renewable energy sources like wind power challenges power grid stability.
• Traditional Under-Frequency Load Shedding (AUFLS) methods struggle with the dynamic nature of modern grids, especially with electric vehicle (EV) participation.
• Maintaining grid frequency within safe limits is critical for reliable power system operation.

Purpose of the Study:

• To develop a robust optimization framework for adaptive Under-Frequency Load Shedding (AUFLS).
• To enhance AUFLS capabilities for managing uncertainties from high wind power penetration and dynamic EV charging.
• To improve power system stability and resilience in the face of renewable energy variability.

Main Methods:

• Proposed a novel bi-level robust optimization framework incorporating an adaptive non-parametric Kernel Density Estimation (AAKDE) for wind power prediction.
• Introduced a strategic EV shedding queue mechanism prioritizing discharge based on real-time state-of-charge and charging behavior.
• Integrated a reinforcement learning model for real-time adjustment of AUFLS decision-making to optimize frequency stabilization.

Main Results:

• Simulations on an upgraded IEEE 39 bus test system demonstrated a reduction in load shedding requirements by over 50% compared to traditional AUFLS.
• The proposed method effectively maintained system frequency within safe operational limits under high renewable variability and EV integration scenarios.
• Achieved superior performance in enhancing grid stability and resilience against dynamic grid conditions.

Conclusions:

• The novel bi-level robust optimization framework significantly improves AUFLS performance for grids with high renewable energy penetration and EV participation.
• Adaptive non-parametric methods and intelligent EV management are key to developing smarter, more resilient power systems.
• The research provides a pathway for advanced AUFLS strategies to ensure grid stability in evolving energy landscapes.