Machine Learning Based Efficient Protection Scheme for AC Microgrid

Abstract

Micro grids have become popular as a way to reduce carbon emissions and use nonrenewable energy sources to produce power. Microgrids allow users to generate and regulate energy as needed, reducing their reliance on the utility grid. They may also sell excess electricity to the grid and make money. Due to its simple design, fast installation, and easy maintenance, photovoltaic systems are a vital microgrid resource. Microgrids threaten the reliability and optimum functioning of major power grids. It's crucial to discover defects early and fix them before catastrophic system breakdown. This research proposes a unique method based on Discrete wavelet transform and ensemble of Decision tree classifier for detecting and classifying microgrid faults. Once the particular fault type is recognised and categorised, a suitable protective strategy may be used to address it early, enhancing the system's overall safety.