Artificial Neural Network Based Intelligent Controller Design for Grid-Tied Inverters of Microgrid under Load Variation and Disturbance

The exponential use of Distributed Generation (DG) resources such as solar, wind, and other renewable energy sources is owing to global climate change and an increase in electric demand. Microgrids comprise many parallel-connected distributed generations (DG) units that use coordinated control techniques and may operate in both grid-connected and islanded modes. In order to integrate renewable energy resources into the electrical grid, three-phase Voltage Source Inverters (VSI) are essential. Traditional linear control approaches used in VSI have a high sensitivity to parameter changes and delayed transient response, resulting in large oscillations in the microgrid frequency and voltage. In recent years, innovative control methods for power converters in microgrids have been successfully developed as replacements for traditional cascaded linear controllers. This paper introduces an Artificial Neural Network (ANN) based controller that compensates for primary frequency and voltage variations faster than the current state-of-the-art methods. The complete converter circuit model is designed and simulated using MATLAB software to obtain a dataset with various load conditions. The ANN-based controller was designed, trained, and tested in a Google collaboration environment. It auto-tunes the model based on the output parameters obtained from MATLAB, improving the system's overall performance. The simulation results suggest that the ANN-based controller outperforms the traditional controller in terms of performance and is more suited for AC Microgrid control.