Human Performance Detection Using Historical Operation Data of Nuclear Power Plants
Dai Xinvu, Ming Hsien Yang, Zhihui Xu, Gu Haixia · 2022
Logs of operator actions provide a valuable resource to review the circumstances of nuclear power plant operations, making them indispensable for investigating past incidents. The traditional operator log review is done manually, which has major problems such as time consumption and low efficiency. This document suggests a system for automated detection, split into three sections: procedure, step, and action. By understanding the fundamental principles and characteristics of the operational protocols, we can create uniform action sequences for comparison and analysis of the operator's activity log to detect potential human errors. The paper explains the method and discusses the advantages and limitations of the proposed method.