Traffic Environment Complexity Identification And Risk Warning Based On Driver's EEG Data

When the driver is in the complex driving scene,the EEG signal is different from that in the ordinary environment.The complex traffic environment factors in the traffic scene usually cause the driver to produce a certain amount of brain load,and then cause the driver's physiological stress response,increasing the probability of traffic accidents.Therefore,it is great significance for driving safety to study the complexity of traffic scenes and analyze the relationship between traffic environment factors and EEG characteristics.At present,the research on driving safety lacks theoretical basis,specifically,the identification of complex traffic factors,complex traffic scenes and early warning is not enough.In this paper,based on the EEG characteristics of drivers under the influence of complex traffic factors,a traffic factor complexity identification model and a traffic factor complexity quantization model based on EEG data of drivers are established.Firstly,based on the existing research on complexity,this paper defines the complexity of traffic environment,classifies the factors affecting the complexity of traffic environment,and introduces the research method of the factors influencing the complexity of traffic based on drivers' EEG signals.Based on the principle of the research method,this paper selects three kinds of dynamic traffic factors: pedestrian crossing,vehicle deceleration in front and vehicle lane change next to each other,and carries out simulation driving experiments on them,laying a data foundation for the research on the correlation between traffic environment complexity and drivers' EEG characteristics.Secondly,based on the EEG signals collected in the experiment,the EEG signals from the event stimulus segment and the normal driving segment were extracted from the time Windows of 3s and 5s respectively for the analysis of EEG indicators.In this paper,the optimal time window and the optimal EEG index are selected by mathematical statistics,and the characteristics of the significant EEG index are analyzed.At the same time,based on the optimal EEG indicator set,this paper USES binary logistic regression analysis to establish the identification models of the factors affecting the traffic complexity of three kinds of dynamic traffic factors,and determines the optimal threshold of the model.Finally,based on the comparison experiments of two kinds of traffic factors with different difficulty,the paper determines the complexity quantization method of different traffic factors.In this paper,the difference between the predicted value and the true value of the wave power spectrum density in the electroencephalon region is selected as an index to quantitatively compare the complexity of traffic factors.At the same time,a driving risk warning mechanism based on identification model threshold and traffic factor complexity is proposed for different traffic factors.