Power Analysis of EEG Data in Multiple Brain Regions during Fatigue Driving

The motor driver fatigue is a major cause of traffic accidents. In this paper, we designed a highway simulated scene by Unity3D software, and collected EEG data from the brain in the simulation experiment. The autoregressive (AR) model spectrum estimation method gives better frequency resolution and is widely used for electroencephalogram (EEG) analysis. The EEG data in different rhythms were studied by AR burg power spectrum method during different phases of simulated driving. The results demonstrated that the relative average power spectrum of the EEG data was significantly increased in the $\delta$ -rhythm of the most brain regions during the post-driving process, and the θ- and α-rhythms were significantly increased of the prefrontal region and the occipital region. At the same time, the EEG indicators R(α/β),R(δ/β), and R((α+β)/δ) were very sensitive to driver fatigue and had significant changes in the post-driving process. With the increase of simulated driving time, R(α/β) and R(α/β) were increase significantly, however, R((α+β)/δ) decreased significantly. This important topic of brain fatigue can provide an objective and effective indicator for studying driving fatigue.