ANN-Based Real-Time Optimal Voltage Control In Islanded AC Microgrids

Abstract

The purpose of this paper is to explore an innovative primary control strategy for a voltage source inverter (VSI). This strategy involves integrating an Artificial Neural Network (ANN) into the proportional resonant (PR) regulator within the inner loop of the primary control system. The primary objective is to regulate the output voltage and minimize deviations under various operating conditions, thereby enhancing the inverter's overall performance. The ANN, employed as a predictive analytic method, facilitates real-time tuning of the PR controller parameters. In addition, the outer level of primary control incorporates a droop control loop to distribute power among distributed generators (DG). This proposed approach reduces the converter control system's reliance on specific operating conditions and seamlessly accommodates varying loading conditions. To ensure adaptability and stability in various operating conditions, real-time simulations are implemented using OPAL-RT (OP4510), and the results demonstrate the efficacy of the proposed control strategy.