A human error classification framework for intelligent nuclear power control centres: a cognitive perspective

Intelligent nuclear power plants integrate advanced information and communication technologies,such as big data, artificial intelligence, and the Internet of Things, enabling self- awareness, selfdiagnosis,and self-optimization. Due to the complexity of nuclear plant systems, operators are still necessary for overseeing the operation of intelligent nuclear power plants. Current classification methods are insufficient to distinguish between different Levels of Automation (LOA) in terms of human errors within intelligent nuclear power plants. This study integrates the four-stage information processing model with the Skill-Rule-Knowledge (SRK) model to form a cognitive-behavioral model for operators. By deepening the connection between operators’ actions and cognition, this approach identifies modes of human error at various LOA. Furthermore, through a case study of the intelligent procedure, this paper analyzes the differences between intelligent and conventional nuclear power plants in terms of tasks, behaviors, cognitive patterns, and potential error modes. The insights gained provide a valuable reference for enhancing the safety and reliability of nuclear power stations.

Xinran Li, Beiyuan Guo,Zhihui Xu.A human error classification framework for intelligent nuclear power control centres: a cognitive perspective.JOURNAL OF NUCLEAR SCIENCE AND TECHNOLOGY,https://doi.org/10.1080/00223131.2025.2521021