CAREER: Robust Data Based Control and Estimation for Resilient DC Microgrids

This NSF CAREER project aims to promote the widespread adoption of complex DC microgrids. While DC distribution networks hold substantial advantages over traditional ac systems in terms of efficiency, their adoption has been gradual due to the lack of research to address the unique protection and stability challenges. This project will bring transformative change by enabling broader implementation of dc microgrids across various scales and industries, ranging from energy efficient buildings and electric vehicle charging stations to aerospace applications. This will be achieved by the development of innovative data-driven methodologies for control, fault and cyber-attack detection, and stability analysis. The intellectual merits of the project include advancing data driven control theory and fault detection and identification, and the development of algorithms to improve the power quality and resiliency of dc microgrids. The broader impacts of the project include research and education tasks for training the next generation of power engineers through new course developments, internship opportunities with industry partners, and efforts to increase the participation of underrepresented students in STEM areas. The expected results have potential applications beyond their immediate scope, extending to areas such as ac power systems, robotics, and vehicle control. This proposal aims to address the current challenges which hinder the broader adoption of DC power systems, including high impedance fault detection, vulnerability to cyber-attacks, and power management issues associated with nonlinear constant power loads. Robust Koopman and behavioral data-based control strategies will be developed for the primary and secondary control of dc microgrids with constant power loads, ensuring stability and power quality. Data driven fault and cyber-attack detection methods will be developed to enhance the resiliency of DC microgrids, revolutionizing traditional model-based protection strategies. In addition, online data driven stability methods will be devised by combining Koopman theory and machine learning techniques. They will provide DC microgrids operators with an accurate knowledge of the stability margins of the network, improving their reliability and robustness. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.