Advanced Power Conversion and Control for Next-Generation Data Center DC Architectures

The rapid expansion of artificial intelligence and digital infrastructure is driving unprecedented growth in data center energy demand. As computing intensity increases, rack power levels are approaching hundreds of kilowatts to megawatt scales, placing significant pressure on conventional low-voltage power distribution architectures. These trends are accelerating the transition toward next-generation data center DC architectures, which aim to improve efficiency, scalability, and power density while enabling flexible and reliable operation.Isolated DC–DC converters, particularly dual-active bridge (DAB) topologies, have emerged as key building blocks in these architectures due to their bidirectional power transfer capability, high power density, and control flexibility. However, several critical challenges remain. Existing modeling approaches often neglect parasitic effects and non-idealities, limiting accuracy across wide operating ranges. Efficiency optimization is constrained by incomplete soft-switching conditions and multi-objective trade-offs. At the system level, coordinated control of multi-module configurations is further complicated by parameter uncertainties, dynamic imbalance, and fast load variations.Recent advances in wide-range modeling, high-efficiency modulation strategies, and data-driven control methods provide new opportunities to address these challenges. In particular, the integration of physics-informed modeling, real-time monitoring, and intelligent control is enabling improved performance, reliability, and scalability of DC power systems in data centers.This Research Topic aims to bring together cutting-edge research on advanced power conversion technologies and system-level control strategies for next-generation data center DC architectures. The focus is on bridging device-level characteristics and system-level performance through accurate modeling, efficient operation, and coordinated control of modular power systems.We welcome Original Research Articles, Review Articles, Mini-Reviews, and Perspectives addressing, but not limited to:• Advanced modeling and characterization of isolated DC–DC converters (e.g., DAB, multi-level topologies) considering parasitic and non-ideal effects• Physics-informed and data-driven approaches for wide-range steady-state and loss modeling• Soft-switching mechanisms and strategies for extending high-efficiency operation across operating regions• Multi-objective optimization of efficiency, power density, and dynamic performance• Parameter design methodologies for cascaded and modular DC architectures• Dynamic balancing and coordinated control of multi-module systems (e.g., ISOP configurations)• Real-time simulation, hardware-in-the-loop (HIL), and experimental validation of data center power converters• AI and machine learning applications in modeling, optimization, and control of data center DC power systemsSubmissions that demonstrate high-efficiency, scalable, and experimentally validated solutions are particularly encouraged. This Research Topic aims to support the development of next-generation high-performance and energy-efficient data center power infrastructures. Topic editors: Jiachen Tian, Nikolay Hinov Source: https://www.frontiersin.org/research-topics/81204/advanced-power-conversion-and-control-for-next-generation-data-center-dc-architectures