AC Microgrid Modeling and Adaptive Control Using Biomimetic Valence Learning: An AI-Based Approach

Abstract

AC microgrids play a crucial role in integrating distributed energy resources and facilitating localized power management in contemporary power networks. Nevertheless, conventional droop control methods in these microgrids have constraints in guaranteeing precise power distribution, stability of voltage/frequency, and flexibility in response to changing operating conditions. This study introduces an approach, with adaptive droop control using Biomimetic Valence Learning (BVLAC). Inspired by the emotional and rational decision-making processes within the brain, BVLAC dynamically adjusts droop coefficients, optimizing power sharing and transient response in microgrid operation. Simulations were conducted using SIMULINK/MATLAB and the results showcase the superiority of the proposed BVLAC approach in achieving precise power-sharing, maintaining voltage and frequency stability, and improving the control performance of microgrids, under varying load conditions. This work advances the field of microgrid control by offering a robust, AI-inspired solution for the challenges faced by conventional droop control techniques.