The ERP characteristics in the process of hazard identification
摘要
The prevention of accidents is aided by having a strong ability to identify hazards. The objective and quantitative assessment of hazard identification ability through event-related potential (ERP) experiments is of significance for person-job safety matching in high-risk positions. In this study, we first designed and conducted an electroencephalogram (EEG) experiment related to the hazard identification process. Subsequently, two indicators reflecting the hazard identification process were extracted from the behavioral data obtained during the experiment: hazard identification speed and hazard identification accuracy. Finally, time-domain and frequency-domain analysis methods were employed to investigate the ERP characteristics and patterns in the hazard identification process. The results showed that: (1) the low and high hazard identification accuracy groups (L-HIA and H-HIA) demonstrated significantly different N100 and P200 components, as well as beta, theta, and alpha power; (2) the fast and slow hazard identification speed groups (F-HIS and S-HIS) demonstrated significantly different N100, P200, and P300 components and beta power; (3) the average power value of theta wave in the central frontal region (Plow < 1.22 µV², Phigh > 1.99 µV²) can be used as the grading standard for hazard identification accuracy; (4) the average peak voltage value of the P300 component in the occipital region (Ufast < 1.78 µV, Uslow > 5.67 µV) can be used as the grading standard for hazard identification speed. An independent validation sample further confirmed the internal reproducibility of these thresholds, achieving 85.7% classification accuracy under the same EEG system and task paradigm. It’s conducive for enterprises and individuals to master the hazard identification ability of employees to train and improve their ability.