Abstract


A new computer simulation approach is proposed for obtaining Human Error Probability (HEP) parameters which are required for Human Reliability Analysis (HRA) as related with Probabilistic Safety Assessment (PSA) for Nuclear Power Plant (NPP). The target HEP parameters are the ones normally represented by Human Cognitive Reliability (HCR) curves. The HCR curves are first obtained from the authors¡Ç previous laboratory experiment and they are analyzed for clarifying probabilistic factors in human cognitive behaviors for diagnosing an abnormal transient of NPP. Then, a human model for simulating such cognitive behaviors is constructed based on a human modeling framework that is expanded from J. Reason¡Çs memory model. By utilizing the human model as Monte Carol simulation to simulate human probabilistic factors, the HCR curves are obtained for diagnosing abnormal transients. The inter-comparison is made between the HCR curves obtained from the experimental data and the ones that obtained by such human model simulation, and its turns out that the both HCR curves agree well with each other. Furthermore, the developed human model of human diagnostic process is implemented into an integrated simulation system of human-machine interaction for a real NPP.

Powered by JDBS Ver.2.1.0
(C) 2020 Hirotake Ishii