Machine Learning Based Techniques for Fault Detection in Power Distribution Grid: A Review
Oladapo Tolulope Ibitoye, Moses Oluwafemi Onibonoje, Joseph O. Dada
Abstract
One of the many issues with the availability of usable power is constant faults in power distribution network (PDN). Fault in distribution network is an anomaly situation that prevents quality power from getting to consumer units when due. Detection of faults based on parametric monitoring using automated machine learning techniques offer tremendous potential. The utilization of machine learning techniques for detecting various fault conditions in power grid is a major solution to power quality problems as it provides a reliable, efficient, and fast approach to resolving issues of power failure. This article presents a brief overview of machine learning methods for fault detection in PDN. The article discusses the pros and cons of existing machine learning methods such as; artificial neural networks, deep learning techniques, support vector machines, k-nearest neighbor, and decision trees. Further research directions toward effective machine learning-enabled systems of fault detection in power grid were also suggested.