Enhancing Smart Grid Security with Machine Learning: A Comparative Study of Supervised and Unsupervised Techniques

The smart grid enables the collection of enormous quantities of high-dimensional, multi-type data on the operation of the electric power grid by integrating modern metering infrastructure, control technologies, and communication technologies. Smart utilities are more vulnerable to cyber-attacks because of their dependence on advanced information and communication technology. Cybersecurity in smart grids has been the focus of numerous studies. This study provides a comprehensive and unique examination of the potential of machine learning to secure a smart grid infrastructure by evaluating supervised and unsupervised machine learning models. This study emphasizes the prospective application and limitations of these models. It evaluates the most recent approaches and considers performance metrics such as computing complexity and accuracy. This study also identifies critical research issues and future approaches for cybersecurity threat identification in smart grids. The three primary sections of this article are a taxonomy of machine learning techniques, a review of related studies, and a concise summary of the results and their significance.