Enhancing AC microgrid control and optimization through advanced inner control loop and artificial intelligence integration

This paper presents a data-driven approach for voltage and current regulation in islanded AC microgrids. The method uses an Intelligent Inner Control Loop (IICL) powered by a Deep Neural Network (DNN). The DNN is trained on 5000 samples obtained from PI controllers under diverse operating conditions, focusing on trajectories that converge to steady state. The network has five hidden layers with sigmoid activation functions and is optimized using the Levenberg-Marquardt (LM) algorithm, while cross-validation ensures reliable generalization to unseen conditions. The proposed controller was validated through Simulation-in-the-Loop (SIL) in MATLAB/Simulink, where it processes dynamic inputs and generates outputs in real-time within the simulation environment. This allows much faster response and eliminates the need for manual tuning, outperforming traditional PI controllers. Simulation results confirm significant improvements. The root mean square error (RMSE) for voltage tracking decreases from 0.2354 to 0.01483. Total Harmonic Distortion (THD) drops from 3.45 % to 0.98 %. Settling time improves from 0.3 s to 0.05 s. Power-sharing accuracy is maintained within 5 %. Additional comparisons highlight the advantage over advanced controllers. For instance, fuzzy logic controllers achieve a THD of 3.23 % and a settling time of 0.1 s, while ANN-based controllers obtain 1.19 % THD and 0.1 s settling time. However, both remain inferior to the proposed DNN, which consistently outperforms them across all key performance indices. The method demonstrates promising precision, adaptability, and robustness under dynamic conditions, while its lightweight architecture supports efficient real-time operation. SIL validation indicates its potential for practical deployment. • Implementation of the ICL in AC Microgrids Using DNN for Voltage and Current Control. • Eliminating Need for Manual Adjustment of Microgrid ICL Control Parameters by IICL. • Faster system stabilization by intelligent controller with fast voltage and current adjustment. • Improved Power Quality through Reduction of Harmonic Distortion by the IICL. • Maintaining Accurate Power Division Among Distributed Generators with the IICL.