2024年3月6日(水) 14:05 〜 15:45 311/312 (Conference Center 3F)

Chair:Minkyu KIM(Korea Atomic Energy Research Institute)
Co-Chair:Michelle BENSI(University of Maryland)

Since the Fukushima Daiichi nuclear power plant accident caused by the 2011 Great East Japan Earthquake, there has been a great deal of interest in the safety of nuclear power plants against external hazards. External hazards can cause direct damage to nuclear power plants, and their timing, location, and severity are highly uncertain and often unpredictable. Therefore, many technologies are being developed worldwide to evaluate and improve the safety of nuclear power plants against external events. In addition, as artificial intelligence-related technologies have been developed quickly in recent years, many technologies utilizing artificial intelligence technology have been developed in engineering and have been proposed for use in conjunction with safety assessments for critical infrastructure such as nuclear power plants. While these advancements come with substantial opportunities to increase efficiency, challenges arise when seeking to use these technologies to model complex systems and processes.
This special session is organized to collect and discuss technologies related to safety evaluation and improvement of nuclear power plants using smart technology. This session will introduce and discuss the current status of engineering technology development using machine learning, smart sensing, and clustering. Machine learning techniques applied to engineering and safety assessment can play a role in efficiently predicting the severity of hazards or the response of structures (e.g., using machine-learning-derived surrogate models that can emulate high-fidelity models with high computational costs). Smart sensing technology can can integrate the response of systems (e.g., the acceleration time history of conventional seismic monitoring sensors) and analyses of structural behavior (e.g., noise filter and response spectrum transformation can be performed directly on the sensor). Additional opportunities existing using clustering technologies. For example, research is underway on how to construct a seismic source model for seismic hazard assessment by utilizing clustering technology using artificial intelligence technology. This session is expected to contribute to the utilization of smart technologies to improve the safety of nuclear power plants, highlighting both opportunities and challenges.