Automatic Classification of Driving Mental Fatigue with Eeg by Wavelet Packet Energy and Kpca-svm

Driving mental fatigue is a main cause of some serious transportation accident and it has drawn increasing attention in recent years. In this study, an automatic measurement of driving mental fatigue based on the Electroencephalographic (EEG) is presented. Fifteen healthy subjects who performed continuous simulated driving task for 90 minutes with EEG monitoring are included in this study. The feature vectors of ten-channel EEG signal on prefrontal, frontal, central, parietal and occipital regions are extracted by wavelet packet transform. Kernel principal component analysis (KPCA) and support vector machines (SVM) are jointly applied to identify two driving mental fatigue states. The results show that wavelet packet energy (WPE) of EEG is strongly correlated with mental fatigue level on prefrontal frontal central and occipital regions. Moreover, the KPCA method is able to effectively reduce the dimensionality of the feature vectors, speed up the convergence in the training of SVM and achieve higher recognition accuracy (98.7%). The KPCA-SVM could be a promising candidate for developing robust automatic mental fatigue detection systems for driving safety.