서강대학교 전자공학과

 Selected for the “2025 Basic Research Laboratory (BRL) Program

A joint research team consisting of Professor Hongseok Kim (Department of Electrical and Electronic Engineering, Sogang University), Professor Seongju Ryu (Department of System Semiconductor Engineering, Sogang University), Professor Youngmin Lee (Department of Artificial Intelligence, Sogang University), and Professor Byungkwon Park (Department of Electrical Engineering, Soongsil University) has been newly selected for the 2025 Basic Research Laboratory Support Program (Advanced Track) funded by the Ministry of Science and ICT and the National Research Foundation of Korea. The research project is titled “A National Power Grid Optimization Laboratory Based on Physics-Informed Neural Networks with Built-in Hardware Accelerators,” and will run for three years from June 1, 2025, to May 31, 2028. The total grant is 1.5 billion KRW (approximately 500 million KRW per year).

Although the share of renewable energy in Korea is growing rapidly, the power grid remains highly vulnerable to unpredictable fluctuations. As solar and wind power installations increase, the need to balance supply and demand in real time becomes more critical. However, traditional methods struggle to handle large-scale problems and often require lengthy computation times, making them difficult to apply in actual operation. Consequently, grid stability deteriorates and the risk of frequent blackouts increases.

To address these societal challenges, this research aims to develop “next-generation power grid optimization technology that combines AI with dedicated hardware chips.” First, AI-based learning models will be used to rapidly assess the grid’s state and propose optimal operational strategies in real time. Simultaneously, a dedicated hardware accelerator chip will be designed to perform complex calculations far more quickly. Through this approach, we will develop a power-grid foundation model capable of predicting and controlling power flows in real time at speeds tens of times faster than existing systems.

Once commercialized, this technology is expected to dramatically reduce grid instability caused by the expansion of renewable energy and significantly lower national power operation costs. It will also prevent blackout incidents, thereby enhancing public safety and industrial productivity, and will contribute to achieving RE100 goals. In the long term, we plan to evolve this solution into a platform that can be used not only by power-sector companies but also by local governments and public institutions.

Professor Hongseok Kim’s Networking for Intelligence Computing and Energy LAB conducts convergent research on AI computing and energy/power-system optimization. Focusing on physics-informed AI, the lab designs AI accelerator hardware and develops power-grid optimization algorithms that span both software and hardware. In particular, it has established its technical expertise by publishing papers continuously in top energy-related journals—such as IEEE Transactions on Power Systems (IEEE TPWRS), IEEE Transactions on Industrial Informatics (IEEE TII), Applied Energy, and Energies—on topics including AI-based power/grid optimization and energy trading.