Survey of Data-Driven Methods for Power Systems

The integration of renewable energy sources and distributed energy resources is transforming the electric power system, creating both opportunities for improved sustainability and challenges related to system complexity and uncertainty. While traditional model-based approaches have historically been used for power system analysis and control, the increasing availability of data from sensors, SCADA systems, and smart meters has spurred the development and application of data-driven methods. This survey paper provides a comprehensive overview of state-of-the-art data-driven techniques applied to power systems, addressing key areas such as predictive analytics for forecasting, state estimation for real-time monitoring, fault detection and diagnosis for enhanced reliability, control and optimization for improved efficiency, and cybersecurity for enhanced grid resilience. These data-driven approaches offer advantages over traditional methods, including data-driven decision making, enhanced system understanding, improved system performance, and adaptation to changing operating conditions. By leveraging data analysis, these methods enable more accurate insights into system behavior, uncover hidden patterns, and optimize operations for improved reliability, efficiency, and security of the modern power grid.