Voltage Stability Assessment of Renewable Microgrids Using Deep Learning

This study investigates the use of artificial neural networks (ANNs) to predict the voltage stability of a renewable microgrid under dynamic load profiles. By integrating photovoltaic (PV) and wind power generation models, the contribution of renewable energy sources (RESs) in enhancing voltage stability can be analyzed. At first, the PV curve of voltage stability will need to be obtained by varying realreactive power of the load to obtain the critical point. Afterwards, Stability Index (SI) values would be calculated from the obtained power and voltage data of the load bus. A Long Short-Term Memory (LSTM)-based neural network model was developed to train and test 70% and 30% of the load profile data, respectively. A detailed SIMULINK simulation model is developed to assess the microgrid stability and reliability. Additionally, it is aimed to minimize the root mean square error (RMSE) of the load data, addressing errors or inconsistencies that may arise during prediction. This work offers insights into the application of machine learning for realtime voltage stability monitoring and prediction which is an essential step for ensuring the stable operation of modern renewable microgrids.