Collaborative Research: Hierarchical Intelligent and Adaptive Techniques to Enable Resilient DC Power Systems

As the adoption of energy sources and loads with inherent dc voltage continues to increase, an electric system based on dc power can offer tremendous advantages over ac, with higher efficiency, less power conversion stages, smaller footprint, and higher reliability. For these reasons, dc power systems and microgrids are now used in electric vehicles, ships, aircraft, and in rural areas. However, electrical faults in dc power networks can lead to extremely dangerous situations which are more difficult to interrupt than their ac counterparts, particularly due to the lack of zero voltage crossings. Moreover, high impedance faults in the form of electrical arcs, such as those caused by loose connections or chafed wires, are very difficult to detect because of the low fault current. The high penetration of electronics loads with advanced controllers make the fault detection and localization even more challenging. To increase the safety and resiliency of dc based systems, the proposed project will address these technical challenges in detecting high impedance faults in dc power systems by developing intelligent and adaptive fault detection, localization, and isolation techniques that are built upon a comprehensive and systematic fault modeling and characterization study. These techniques can significantly improve the performance of existing and future dc systems to enable their wide adoption at larger scales, which can provide efficient and reliable interfaces to many renewable resources, energy storage units, and modern electronic loads and align with the nation's initiatives in using clean and green energy. This project is intrinsically multidisciplinary by bringing advanced and exciting modern control theories, artificial intelligence, and signal processing techniques into electric power engineering. The tasks in this project involve a wide range of expertise and experience from software simulation and control algorithms to hardware testing; from circuit level study to system level implementation, which provides a unique and high quality training opportunity for future engineers. The proposed educational activities will also broaden participation of women and other under-represented students. The goal of the proposed research is to develop fault detection, localization, and isolation techniques for modern dc power systems through a hierarchical approach with intelligent and adaptive functionalities. It addresses the most challenging issues in the protection of dc power systems with a systematic and transformative effort. The fault modeling and characterization study of the proposed project will generate fundamental and critical knowledge of high impedance faults in modern application settings through comprehensive experimental and analytical approaches. The proposed high impedance fault detection and localization techniques will take into account the effect of advanced controllers through dynamic parameter estimation. The adaptive and integrated fault detection and localization schemes to be developed will significantly enhance the existing protection system design and online stability assessment methodologies by adopting modern nonlinear control theory and artificial intelligence tools. The proposed research is expected to produce significant results of both theoretical and practical values to the field of dc power systems. When successfully completed, the project has the potential to revolutionize the control and protection aspects of dc power systems, minimizing the adverse impact of high impedance faults and constant power loads. The proposed techniques can be applied to dc systems in different scales ranging from isolated dc distribution networks to interconnected dc microgrids, to improve the fault protection effectiveness and therefore their resiliency. 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.