Research On Channel Selection Method Of Eeg And Brain Region Analysis For Fatigue Driving

Fatigued driving is a significant contributor to traffic accidents.There are some issues with common EEG data of 32 channels,64 channels and 128 channels,such as difficult acquisition,high data redundancy and difficult practical application.This thesis develops a channel selection model based on ReliefF＿SFS to address the problem of how to reduce the number of channels while maintaining classification accuracy,as well as the current lack of brain region analysis for fatigue driving.At the same time,a fatigue brain area analysis model based on functional brain network is developed in order to further validate the effectiveness of channel selection and to understand the neural mechanism of fatigue.The following are the main research topics and milestones for the subject:(1)Research on the extraction of various features and fusion features for channel selection.The features of real EEG data are retrieved from three aspects: time domain,frequency domain and entropy,based on the diverse recognition abilities of distinct features on EEG signals.To give a feasible feature combination approach for future research,the features with better performance are fused,and the recognition ability of single and fused features is compared.(2)EEG signal channel selection approach for fatigued driving research.A new channel selection method called ReliefF＿SFS is proposed to address the problems of data redundancy and acquisition difficulty in the practical application of full channel data.It combines the ReliefF algorithm and the sequential forward selection(SFS)algorithm.When only T6,O1,Oz,T4,P3 and FC3 are employed,the classification accuracy under Theta＿Std+FE combined with ReliefF＿SFS achieves 99.45%.The strategy suggested in this thesis not only ensures the recognition accuracy,but also reduces the number of channels when compared to other models based on the same data set.(3)A functional brain network-based investigation of fatigue brain regions.After extracting diverse features from the genuine EEG data set,the functional brain networks in resting and fatigue states are built using the network sparsity method.By calculating the network measures of functional brain network,drawing brain topographic maps,and analyzing the changes in the brain from resting to fatigue states,the important brain regions that produce driving fatigue can be identified,the effectiveness of channel selection can be further verified,and support for the analysis of fatigue neural mechanisms can be provided.The experimental findings suggest that the upper midline region,temporal region,parietal region,posterior temporal region and occipital region are the brain regions most associated with the generation of driving fatigue,which are largely congruent with the results of channel selection.