Machine Learning: A Comprehensive Exploration of Fault Detection and Diagnosis in Smart Grids
Peter Onu, Charles Mbohwa, Anup Pradhan
Abstract
The increasing complexity of smart grids, driven by the integration of renewable energy sources and advanced technologies, presents new challenges for Fault Detection and Diagnosis (FDD). As a promising solution, machine learning (ML) techniques have emerged to address these challenges. This comprehensive article provides an in-depth review of ML techniques utilized for FDD in smart grids, offering a broad overview of the existing literature on the advantages and limitations of various methodologies such as artificial neural networks, support vector machines, decision trees, and deep learning. The paper also explores vital aspects like data pre-processing, model evaluation and validation. Additionally, it elucidates the potential of Model Predictive Control (MPC) and Fuzzy Logic Control (FLC) tools for automation and systems control within smart grids, strengthening the connection between FDD and overall grid management. Practical case studies are presented to illustrate the effective application of these techniques. Serving as a valuable starting point for researchers, this article expands knowledge on machine learning techniques for FDD in smart grids. It contributes new insights and advancements to the field, playing a critical role in the development of a more reliable and resilient power grid.