This paper presents a formal, research-oriented model designed to support a broad range of artificial intelligence and machine learning (AI/ML) tasks related to managing the life cycle of nuclear power plants (NPPs). The model consists of a hierarchical graph structure, a semantic layer, and a classification system. It also includes links to a corpus and associated metrics. It is intended primarily to advance research on AI/ML methods that support NPP life cycle management, including engineering, construction, commissioning and startup, and decommissioning. At the same time, with appropriate adaptation and optimization of selected components and algorithms, the model can be adapted for deployment in real NPP projects. The model does not have to be implemented in full and can be adapted, simplified or extended to meet the specific objectives of AI/ML applications for implementation in NPP projects. Certain findings and recommendations presented herein may necessitate additional investigation, depending on the specific research objectives or on how the model is implemented for particular applications. While this work concentrates on the life cycle of nuclear power plants, the proposed principles and methods are transferable to other complex systems.
Yuriy Dmitrishin (Sun,) studied this question.