Frequency control of the islanded microgrid including energy storage using soft computing

Area of Science:

• Electrical Engineering
• Control Systems
• Renewable Energy Systems

Background:

• Microgrids are increasingly complex due to distributed generation.
• Conventional controllers struggle with non-linearity and varying operating points.
• Effective frequency control is crucial for stable islanded microgrid operation.

Purpose of the Study:

• To develop an adaptive frequency controller for islanded microgrids.
• To improve the performance of proportional-integral (PI) controllers in dynamic microgrid environments.
• To address limitations of traditional artificial neural networks (ANNs) in controller tuning.

Main Methods:

• A hybrid soft computing approach combining genetic algorithms (GA) and artificial neural networks (ANN) was employed.
• The GA-ANN hybrid optimizes PI controller coefficients for secondary frequency control.
• Online training of the ANN was implemented to adapt PI coefficients without extensive data, avoiding local minima.

Main Results:

• The proposed GA-ANN-based PI controller effectively managed microgrid frequency during islanded operation.
• The controller demonstrated robust performance under various perturbations and system nonlinearities.
• Simulation results confirmed the successful adaptation and optimization of PI control coefficients.

Conclusions:

• The developed self-tuning PI controller offers a significant improvement for microgrid frequency regulation.
• The GA-ANN approach provides an effective method for adaptive control in complex power systems.
• This strategy enhances microgrid stability and resilience, particularly with diverse energy resources.