Design and implementation of model-assisted reduced-order ADRC for power system load frequency control problem with...

Area of Science:

• Control Systems Engineering
• Power Systems Stability
• Nonlinear Control Theory

Background:

• Load Frequency Control (LFC) is critical for power system stability.
• Challenges include communication delays, actuator nonlinearities, and parameter uncertainties.
• Existing methods like PID/FOPID/PI/H∞ controllers struggle with these complex dynamics.

Purpose of the Study:

• To develop a robust LFC strategy using a Model Assisted Reduced-order Active Disturbance Rejection Controller (MRADRC).
• To enhance observer accuracy and controller performance under uncertain and delayed conditions.
• To validate the proposed MRADRC against conventional controllers in realistic power system scenarios.

Main Methods:

• Design of a reduced-order Extended State Observer (ESO) for delay-aware state estimation.
• A two-stage controller tuning approach utilizing the Walrus multi-objective optimizer and multi-criteria decision-making.
• Implementation and testing on single-area LFC plants and the IEEE 39-bus New England system.

Main Results:

• MRADRC demonstrated significant improvements in reducing frequency deviations and peak frequency errors.
• The controller maintained the desired robustness level (2≤ϵ≤5) and achieved a good delay margin.
• Superior performance compared to PID, FOPID, PI, and H∞ controllers was observed across various scenarios.

Conclusions:

• The proposed MRADRC effectively addresses LFC challenges posed by communication delays, nonlinearities, and uncertainties.
• MRADRC offers a robust and high-performance solution for modern power grid control.
• The developed tuning methodology ensures optimal controller and observer bandwidth selection.