Microgrid protection using FACTS controller and machine learning

This research presents a method of machine learning approach for microgrid protection using FACTS Controllers. A Microgrid connected to the Distributed Generators connected with different types of load is very prone to faults. This work uses wavelet transform to predict the failure signals, which is then fed into a machine learning model for real-time identification. These current signals from both ends of the transmission lines are analyzed using wavelet-based multi-resolution analysis, and then compare the results to preset thresholds to generate fault indices. This approach offers a power quality enhanced microgrid protection solution with lower losses and increased dependability.