人工智慧技術應用於放射性廢棄物及其設施之管理評估研究

The deep geological repository (DGR) is recognized as one of the most reliable methods for the long-term disposal of spent nuclear fuel (SFD)/ high-level radioactive waste (HLW). The DGR consists of engineering barriers and natural barriers where coupled thermo-hydro-mechanical-chemical (THMC) processes will occur. Traditional numerical tools are widely used to understand THMC processes but they are quite computationally expensive. Most recently, the application of artificial intelligence (AI) techniques like data-driven machine learning (DDML) or physics-inspiring machine learning approach for HLW is of emerging interest. This project is thus designed to systematically elaborate, analyze and summarize the application of AI technique in the radioactive waste management (RWM) and its associated facilities. Four major tasks will be executed in this project. Task 1 will collect international experience of AI application for RWM. Task 2 will analyze the application of AI for RWM and its associated facility. Task 3 will develop a preliminary AI model for radioactive waste management. Task 4 will deliver the suitable concept and strategy for application of AI in the RWM in Taiwan.