Identification Model For Driving Sustained Attention Level Based On Eeg

The driving sustained attention level affects driving safety is the key factor,then there is a direct relationship between the changes of EEG signals and the sustained attention level,so in order to recognize the driving sustained attention level effectively,an identification method for sustained attention level was proposed based on the signal of EEG,which is the recognizing feature index.Firstly,in order to divide the grade of the sustained attention level,taking the driving reaction time as the index of objective division,a dividing method for sustained attention level was proposed.The efficient screening algorithm is used to the preferential selection of EEG feature parameters,the EEG feature parameters are selected as the recognizing feature index of driving sustained attention level.Finally,particle swarm optimization(PSO)is used to optimize the parameters of the support vector machine(SVM),an identification model for identifying driving sustained attention level is founded.The driving sustained attention level test was designed in the driving simulation bench.The identification model was tested,which is based on the EEG data from the driving simulation test,the main conclusions are as follows:(1)The driving reaction time is used to divide the driving sustained attention level,the difference of driving behavior performance at different grade is comparative analysis,and the result shows that there is significantly difference of the driving behavior performance at different grade,it is verified the rationality of the division method.(2)The Kruskal-Wallis test and Relief algorithm is used to the preferential selection of EEG feature parameters,the difference of the synthetic parameters(a+?)/? of FP1 electrode,?/? of P7 electrode,(?+?)/? of CZ electrode,(a+?)/? of P8 electrode,(a+?)/? of FZ electrode in the different grade of driving sustained attention level are the most significant,they can be used as the recognizing feature index of driving sustained attention level(3)The average accuracy rate of model is 92.19%,the model and method proposed is feasibility and applicability,and it is applicable to identification of driving sustained attention level.